ISSN 0038-3872 II ■JM SOUTHERN CALIFORNIA ACADEMY OF SCIENCES BULLETIN Volume 1 12 Number 1 A112(l) 1-48 (2013) April 2013 Southern California Academy of Sciences Founded 6 November 1891, incorporated 17 May 1907 © Southern California Academy of Sciences, 20 1 3 2012-2013 OFFICERS Julianne Kalman Passarelli, President Bengt Allen, Vice-President Edith Read, Recording Secretary Daniel Guthrie, Corresponding Secretary Ann Dalkey, Treasurer Daniel J. Pondella II, Editor Larry G. Allen, Editor ADVISORY COUNCIL Jonathan Baskin, Past President John Roberts, Past President Robert Grove, Past President John H. Dorsey, Past President Ralph Appy, Past President 2010-2013 BOARD OF DIRECTORS 2011-2014 Lisa Babiloma Brad R. Blood Ann Dalkey Julianne Passarelli Edith Read David Ginsberg Gordon Hendler Andrea Murray Dan Guthrie Gloria Takahashi 2012-2015 Bengt Allen Shelly Moore Ann Bull Dan Cooper Mark Helvey Membership is open to scholars in the fields of natural and social sciences, and to any person interested in the advancement of science. Dues for membership, changes of address, and requests for missing numbers lost in shipment should be addressed to: Southern California Academy of Sciences, the Natural History Museum of Los Angeles County, Exposition Park, Los Angeles, California 90007-4000. Professional Members $60.00 Student Members 30.00 Memberships in other categories are available on request. Fellows: Elected by the Board of Directors for meritorious services. The Bulletin is published three times each year by the Academy. Submissions of manuscripts for publication and j associated guidelines is at SCASBULLETIN.ORG. All other communications should be addressed to the Southern [ California Academy of Sciences in care of the Natural History Museum of Los Angeles County, Exposition Park, Los Angeles, California 90007-4000. Date of this issue 1 1 April 2013 ® This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Bull Southern California Acad. Sci. 112(1), 2013, pp. 1-18 © Southern California Academy of Sciences, 2013 Growth, Development, and Reproduction in Gulf Corvina ( Cynoscion othonopterus) Katie E. Gherard,' Brad E. Erisman,2 Octavio Aburto-Oropeza,2 Kirsten Rowell,3 and Larry G. Allen4 1 Louisiana Department of Wildlife and Fisheries, New Orleans, Louisiana 70118, USA, katiegherard@gmail. com 2Scripps Institution of Oceanography, University of California San Diego, La Jolla, California 92093-0202, USA 3 Department of Biology, University of Washington, Seattle, WA 98195-1800 4 Department of Biology, California State University, Northridge, California 91330-8303, USA Abstract. — Gulf corvina, Cynoscion othonopterus , is a vital component of com- mercial fisheries in the northern Gulf of California, but a lack of information on life history parameters have thus far prevented a comprehensive stock assessment. In this project, 530 specimens of Gulf corvina were collected from commercial gill net fisheries in the Colorado River Delta region in Sonora, Mexico, to characterize population structure, age and growth patterns, age and size at sexual maturity and batch fecundity. Fish ranged from 145 mm to 1013 mm in total length and from 1 to 8 years of age. Von Bertalanffy growth model parameters were: Lo==1006 mm, k = 0.255/yr, tn = 0.616 years. Growth rates of Gulf corvina did not differ significantly between sexes, although females were predicted to reach a larger asymptotic length. Mean size (Lm50) and age (Am50) at sexual maturity from histological analyses of gonad tissues was 294.7 mm and 2.3 years for females and 267.5 mm and 2.0 years for males. Maturity estimates from otolith analyses did not differ between sexes and were similar to maturity estimates derived from gonadal histology, indicating that energy allocation shifts from growth to maturation and reproduction after year two. Batch fecundity ranged from 240,394 to 1,219,342 eggs with a mean of 684,293 eggs per spawn, and was correlated to both total length and gonad-free body weight. The distribution of oocyte diameters and oocyte stages indicate that Gulf corvina is a multiple batch spawner with asynchronous oocyte development and indeterminate annual fecundity. Introduction The Gulf corvina, Cynoscion othonopterus (Jordan & Gilbert 1882), is a member of the family Sciaenidae and is endemic to the northern Gulf of California, the region north of latitude 28°61' S (Robertson and Allen 2008). Adult Gulf corvina are carnivorous, feeding on benthic crustaceans, mollusks, and schooling fishes such as anchovies and sardines (Roman-Rodriguez 2000). Spawning is restricted to the upper reach of the Gulf of California and is correlated to tidal patterns: adults migrate to the Colorado River Delta during the weeks preceding the new and full moons of March and April and form massive spawning aggregations in the estuary (Erisman et al. 2012). The spawning season SOUTHERN CALIFORNIA ACADEMY OF SCIENCES is tightly linked to the timing of historic spring floods of the Colorado River, which created an extensive brackish nursery habitat for Gulf corvina and other fish (Rowell et al. 2005, 2008). The commercial fishery for Gulf corvina is economically significant in the northern Gulf of California (Paredes et al. 2010; Rodriguez-Quiroz et al. 2010) and has a direct conservation impact, because it is centered inside the upper Gulf of California and Colorado River Delta Biosphere Reserve (DOF 2007). The fishery primarily targets the spring spawning aggregations, in which small boats (8-9 nr in length) use gillnets with a mesh size of 14.6 cm and lengths up to 293 m to harvest fish as they move to the river’s mouth to spawn (Roman-Rodriguez 2000; DOF 2007; Paredes et al. 2010; Erisman et al. 2012). Over the past two decades, the Gulf corvina fishery has emerged as an important economic component of the communities of El Golfo de Santa Clara, El Zanjon, and San Felipe in the northern Gulf of California (Roman-Rodriguez 2000; Rodriguez- Quiroz et al. 2010). In El Golfo de Santa Clara, annual harvest increased from 3.2 tons to 1,278 tons between 1993 and 1996 and has ranged between 1,767 to 4,370 tons since 2000 (CONAPESCA 2010). The estimated average annual value of the Gulf corvina catch from 1995 to 2007 in El Golfo de Santa Clara was $2,318,303 USD, constituting about 30% of the total value of all major fisheries in the town (Rodriguez-Quiroz et al. 2010). The Gulf corvina is one of only a few fish species in Mexico that is regulated by an official management plan (DOF 2007), and concerns exist that the stock is overexploited and highly susceptible to a collapse (Musick et al. 2000; Rodriguez-Quiroz et al. 2010; Erisman et al. 2010a). Efforts to conserve the species and to create a sustainable fishery have been developed by government agencies in cooperation with stakeholders and regional non-governmental organizations (DOF 2005, 2007; Paredes et al. 2010). Similarly, important research on the biology and fishery of Gulf corvina has been completed (e.g., Roman-Rodriguez 2000; Campoy and Roman-Rodriguez 2002; Rowell et al. 2005; Erisman et al. 2012). Nevertheless, significant gaps exist related to the life history of Gulf corvina, which impede the completion of a comprehensive stock assessment necessary to regulate harvest. Such information is crucial for fisheries management, since life history traits represent key parameters used in stock assessments and comprise the core determinants of stock responses to both environmental and anthropogenic influences (King and McFarlane 2003; Shin et al. 2005). In this study, we analyzed samples of Gulf corvina harvested by commercial fishers to characterize several life history parameters necessary for a stock assessment; population demographics, somatic and otolith growth, gonad development, and fecundity. Our specific objectives were to characterize population structure, determine growth rate by sex, estimate and validate age and size at maturity using gonadal histology and otolith annuli measurements, and describe variations in batch fecundity in relation to fish length and body mass. Materials and Methods Collection of Specimens A total of 530 Gulf corvina were obtained from commercial fishers at El Golfo de Santa Clara and El Zanjon from March 2009 to May 2011. Four hundred large individuals with a mean total length (TL) of 716 mm were collected from the corvina fishery in March to April in 2009 and 2010. Given the size selectivity of the Gulf corvina gill net fishery (i.e., 14.6 cm mesh size selects for fish larger than 50 cm), an additional 130 GROWTH AND REPRODUCTION OF C. OTHONOPTERUS 3 small specimens (mean = 301 mm) were collected from the bycatch of two other commercial gill net fisheries: the commercial blue shrimp fishery (Litopenaeus stylirostris\ 5-6 cm mesh; September-December 2009) and the bigeye croaker fishery ( Micropogonias megalops ; 6-9 cm mesh; April-May 2011) in order to obtain individuals representing younger age classes. Meristic (e.g., 23-27 soft rays on second dorsal fin) and morphometric (e.g., presence of scaly sheath covering at least half the soft rays on the second dorsal fin) characteristics, known to be diagnostic of Gulf corvina, (Chao 2003; Robertson and Allen 2008) were used to confirm each specimen as C. othonopterus and to avoid inclusion of congeneric species known to inhabit the region ( C. parvipinnis , C. reticulatus, and C. xanthulus) in the study. Total length (TL; 1 mm), total body weight (TW; 0.1 g), and sex were recorded for each fish. Otoliths were removed, dried, and stored, and gonads were removed, weighed (GW; 0.1 g), and preserved in a 10% formalin-seawater solution. Gonad-free weight (GFW) was calculated for each sample as follows: GFW = TW-GW. (1) Age, Growth, and Population Structure Sagittal otoliths (Figure 1) were removed from 492 fish (219 males and 273 females) using the method described by Craig et al. (1999) to analyze patterns related to age and growth. In the laboratory, the otolith weight (0.0001 g) and otolith radius (0.0001 mm) were measured. These data were used for age validations under the assumption that otolith weight and radius increase as fish grow and age and to verify precision of ageing methods (Cailliet et al. 1996). Otoliths were mounted on wood blocks with cyanoacrylate adhesive and a 0.5 mm dorsal-ventral cross-section was made through the focus using a Buehler-IsoMet double bladed low speed saw with diamond edged blades (Allen et al. 1995). Sections were polished using silicon carbide lapping paper and imaged while submerged in water in a black backed watch glass under a dissection scope. Each otolith was aged by two independent readers from digital images of otolith cross sections using ImageJ (Rasband 1997-2009) (Figure 1), as preliminary estimates using direct observations through the scope showed no evidence of distorting the band pattern and did not affect age estimates. The limited seasonal availability of samples across multiple age classes prevented a marginal increment analysis to validate age determination. Flowever, nonlinear regression was used to test for a relationship between age and otolith weight, age and otolith radius, and length and otolith radius to provide some validation for our methods of age determination and to allow for calculations of fish length directly from otolith measurements in future studies. Age and length were designated as the explanatory variables, whereas otolith weight and otolith radius were designated as the response variables. The age and length data were then fit to the following von Bertalanffy growth model (VBGM): Lt = L00[l-e-K(t-to )] (2) where Lt = length at age t, L* = predicted maximum length, K = growth coefficient, and tn = theoretical age at zero length. This equation was fitted for males and females separately, and an analysis of the residual sum of squares (ARSS) was used to compare VBGM’s by sex (Ratkowsky 1983; Chen et al. 1992). An ANCOVA was used to test for differences in length at age by sex. 4 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Fig. 1. Sagittal otoliths of Gulf corvina. (top) Drawing of a whole otolith from an average age 5 yr individual, where the dashed line represents where a section was taken through the focus, (bottom) Transverse section of a sagittal otolith from a 749 mm, age 5 yr individual, where numbers demarcate annuli, core is marked, and there is an example of where interannular width measurements were taken. Reproductive Development and Sexual Maturity Standard histological techniques (Humason 1972) were used to perform microscopic examinations of gonad tissues, which provided a means to distinguish sexually mature individuals from immature individuals that had never spawned before. Briefly, preserved samples of gonad tissues were taken from the central portion of one lobe, embedded in paraffin, sectioned transversely to 5-6 micron thickness, and stained with Haemotoxylin and Eosin Y. Classification of gonadal development stages followed Brown-Peterson et al. (2007). Sexually mature females were classified as containing oocytes in early to advanced (i.e., hydrated) stages of vitellogenesis, whereas sexually mature males were identified by the presence of spermatozoa within spermatocysts, lobules, or fully developed sperm ducts. Diagnosis of sexual pattern followed criteria outlined for fishes GROWTH AND REPRODUCTION OF C OTHONOPTERUS 5 by Sadovy and Shapiro (1987) and Sadovy De Mitcheson and Liu (2008). For both sexes, the mean size at sexual maturity (L,„50) was estimated by fitting a logistic function to the proportion (P,„) of mature fish in 20 mm (L5) size categories: where L Smui is the midpoint of the Lj class, L„,5n is the mean Ls at sexual maturity and r is a constant that increases in value with the steepness of the maturation schedule (Erisman et al. 2010b). The same procedure was used to estimate mean age at sexual maturity (A,„5o) using the proportion of mature fish in each age class. Variations in the widths between annular bands (i.e., interannular widths) in otoliths were analyzed as another estimate of age at sexual maturity to compare with traditional estimates derived from histological analyses of gonad tissues. Rationale for this method was based on two assumptions: (1) otolith growth is generally proportional to somatic growth (Pannella 1971); (2) the ontogenetic shift to sexual maturation coincides with slowed somatic growth as energy is reallocated from growth to reproduction (Charnov et al. 2001), which is reflected in otolith growth rates. Growth rates subsequent to the first year continue to decrease as fish grow at increasingly slower rates. We hypothesize that when sexual maturity is reached, otolith growth rates slow (i.e., smaller interannular widths) to a continuous rate. One hundred and twenty fish estimated to be 5 years of age or older were randomly selected, distances between annular bands (interannular width) were measured from digital images of otolith cross sections (Figure 1 ), and differences between adjacent interannular widths were compared via an ANOVA with a post-hoc Tukey test. Batch Fecundity Batch fecundity (BF), or the number of oocytes released in a single spawning event, was estimated by the gravimetric hydrated oocyte method (Hunter et al. 1985). We analyzed 58 sexually mature females (2009, n = 36 and 2010, n = 22) collected from the corvina fishery with ovaries containing hydrated oocytes. Females with ovaries containing both hydrated oocytes and new (< 6 hr) post-ovulatory follicles (POF’s) were excluded. The presence of new POF’s indicates that spawning may have begun prior to sampling, which could lead to an underestimation of BF (Hunter and Macewicz 1985). Each ovary was first blotted dry with bibulous paper, and three subsamples were then removed from each the anterior, middle, and posterior region of ten ovaries (right or left was chosen at random) and weighed (0.000 lg). Oocytes from each subsample were teased apart from follicles and tissue using forceps and a spatula, arranged in a single layer on a slide using water and a spatula, and imaged. Oocytes were categorized by developmental stage (Figure 2, Table 1) and enumerated using ImageJ (Rasband 1997-2009). Batch fecundity was extrapolated for each sub-sample: where BF = batch fecundity, NOSS = number of oocytes in subsample, SSW = subsample weight (range = 0.0316 to 0.1420 g), and GW = gonad weight, and averaged for each region to verify that hydrated oocytes were distributed uniformly within the ovary. A one-way ANOVA was used to test for differences in BF among the three regions of the ovary. Notably, 25 ovary samples were weighed both before and after formalin preservation in order to estimate changes in gonad weight due to the preservation process and calculate a correction factor. (3) BF = NOSS/SSW x GW (4) 6 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Fig. 2. Oocyte developmental stage classifications of Gulf corvina. (top) SI = stage 1 and S2 = stage 2 under (left) dissection scope and (right) histological section, (middle) S3 = stage 3 under (left) dissection scope and (right) histological section, (bottom) S4 = stage 4 under (left) dissection scope and (right) histological section. In order to verify t he delineation of developmental stages, the vertical and horizontal diameter of up to 55 (mean = 28) oocytes per stage were measured (0.0001 mm) and averaged in 10 randomly chosen subsamples (Macewicz and Hunter 1994) using ImageJ (Rasband 1997-2009). A one-way ANOVA and post-hoc Scheffe test were used to compare oocyte diameters between stages and validate oocyte stage classification (Hunter et al. 1989). An ordinary least-squares regression was used to fit batch fecundity to GFW and length for each year and for the combined data (Hunter et al. 1 985). Gonad-free weight was GROWTH AND REPRODUCTION OF C OTHONOPTERUS 1 Table 1. Description of developmental oocyte stages used to classify Gulf corvina oocytes in fecundity estimates and histological analyses. Stage Description 1 Previtellogenic to early vitellogenesis, oocyte very small (<0.7 mm), uniformly opaque, usually found with stage 3 and 4 oocytes but not as part of same batch 2 Mid to late vitellogenesis, many small oil droplets present, resulting in granular appearance, mean diameter of 0.93 mm 3 Migratory nucleus, oil droplets have coalesced into single droplet, oocyte retains spherical shape, mean diameter of 1.21 mm 4 Hydrated, oocyte becomes irregular in shape and is uniformly translucent, mean diameter of 1.16 mm used, because hydrated oocytes significantly affect the TW of a female (Hunter et al. 1985). ANCOVA was used to compare BF at length and weight between years. Results Age, Growth, and Population Structure Fish ranged in length from 145 to 1013 mm (mean = 605 ± 190 mm) and age from 1 to 8 yrs (mean = 4.5 ± 1 .6 yrs) (Figure 3). Males ranged in length (± SD) from 255 to 895 mm (mean = 636. 1± 145 mm) and age from 2 to 8 yrs (mean = 4.7±1.3 mm), whereas females ranged in length from 215 to 1013 mm (mean = 642± 1 68 mm) and age from 2 to 8 yrs (mean =4.9 ±1.4 mm). No significant relationship was found between either mean length (ANOVA, df = 1, F-ratio = 0.367, p = 0.545) or mean age (ANOVA, df = 1, F-ratio = 1.610, p = 0.205) of males and females. A significant relationship was found between otolith radius and age (n = 488, p < 0.01), otolith weight and age (n = 457, p < 0.01), and between otolith radius and length (n = 488, p < 0.01) (Figure 4). The VBGM equation was estimated for all samples (Lt = 1006[ 1 _e— 0 255 20 -a ;s 0 ■ Juveniles □ Males □ Females laJ r-n, -r£L Fig. 3. Size (top) and age (bottom) distribution of Gulf corvina sampled from the targeted Gulf corvina fishery and as bycatch from other commercial fisheries from 2009 to 2011. in otoliths indicated that at the onset of maturity adjacent interannular widths were significantly different from one another (ANOVA, n = 527, df = 6, 520, F-ratio = 86.691, p < 0.001) (Figure 7). A post-hoc Tukey test indicated that growth rates slowed to a constant rate after age two, as indicated by the comparison of the year 3 - year 4 difference to the year 4 - year 5 difference (p = 0.859). The same results were found when males and females were analyzed separately (p = 0.936 and p = 0.978, respectively). Batch Fecundity The smallest mature female found with hydrated oocytes was 4 yrs of age and 665 mm long. Batch fecundity estimates among the three ovarian regions (anterior, middle, and posterior) were not significantly different (ANOVA, n — 30, df = 18, F-ratio = 0.3384, p = 0.718). Therefore, the distribution of oocyte stages within an ovary were considered to be homogenous, and subsamples thereafter were taken from any region of an ovary. Mean oocyte diameters differed significantly between stages, (n = 762, df =758, F-ratio = 1606.316, p < 0.001), and a post-hoc Scheffe test indicated that all stages were significantly different from one another (p < 0.001 ) except for stage 3 and 4 oocytes (p = 0.053) (Figure 8). Stage I oocytes had a mean diameter of 0.4377±0.0085 mm. Stage 2 oocytes had a mean diameter of 0.9300±0.0095 mm. Stage 3 oocytes had a mean GROWTH AND REPRODUCTION OF C. OTHONOPTERUS 9 3.5 3.0 § 2.5 1 2.0 ■a a c L5 2 1.0 D 0.5 0.0 y = 0.789 ln(x) + 0.781 R2 = 0.699 f I I * i 0 5.0 4.5 - _ 4.0 3 3.5 H ljb 3.0 ’<3 £ 2.5 i I 2-° 5 1-5 - 1.0 - 0.5 0.0 0 3.5 3.0 I 2.5 Age (yrs) y = 0.151 (x1491) R2 = 0.799 , 1 1 ■i A ir Age (yrs) -o C3 Di 2.0 ..5 2 1.0 O 0.5 0.0 y = 0.855 ln(x) -3.536 R2 = 0.678 & 100 200 300 400 500 600 700 Total Length (mm) 800 900 1000 1100 Fig. 4. Plot of nonlinear regressions of: (top) otolith radius and age (n = 488, p < 0.01); (middle) otolith weight and age (n = 457, p < 0.01); and (bottom) otolith radius and TL (n = 488, p < 0.01). Data points represent observed data while dashed lines indicate predicted values. diameter of 1.205±0.014 mm, and stage 4 oocytes had a mean diameter of 1.162±0.007 mm. Stages 3 and 4 were both counted for estimates of batch fecundity. Estimates of BF ranged from 240,394 to 1,219,342 eggs with a mean of 684,293 eggs per spawn. In 2009 samples, a significant relationship was found between BF and both length 10 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Fig. 5. (top) Age at length data by source and the von Bertalanffy growth model calculated from 492 samples of Gulf corvina. (bottom) Plot of residual values by source for length at age (observed-predicted) data included in the growth model. (F-ratio = 26.354, p < 0.001) and GFW (F-ratio 22.812, p < 0.001) (Figure 9). However, in 2010, BF was not significantly related to TL (F-ratio = 1 .333 p = 0.262) or GFW (F-ratio = 3.359, p = 0.082). For the combined 2009 and 2010 data BF was again significantly dependent on both length (F-ratio = 19.044, p < 0.001 ) and GFW (F-ratio = 17.102, p < 0.001 ). The interaction term of the ANCOVA comparing BF and length between years was not significant (df = 54, F-ratio = 2.418, p = 0.126) but the subsequent analysis without the interaction term indicated that BF was significantly correlated to year (df = 55, F-ratio = 34.339, p < 0.001) and length (df = 55, F-ratio = 22.222, p < 0.001). The interaction term of the ANCOVA comparing BF and GFW between years was not significant (df = 54, F- ratio = 2.668, p = 0. 108) but the subsequent analysis without the interaction term indicated that BF was significantly correlated to year (df = 55, F-ratio = 36.538, p < 0.001) and GFW (df = 55, F-ratio = 22.075, p < 0.001). Discussion The collective results of this study indicate that the Gulf corvina is a fast growing sciaenid fish that attains sexual maturity at a relatively small size (200 mm) and young age (2 yrs). With respect to reproductive pattern, the species is best described as a GROWTH AND REPRODUCTION OF C OTHONOPTERUS 1 0.8 0.6 0.4 0.2 0 * / ' / 1 / 1 / 1 / 1 -e — ^ --G — — — — • Expected O Observed A50 (A -2.05) Pm = ( 1 + e -r Smid ) - 1 012345678 Fig. 6. Cumulative length (left) and age (right) frequency plots for male (top) and female (bottom) Gulf corvina sampled from the commercial harvest between 2009 and 2011. gonochoric, multiple batch spawner (asynchronous oocyte development) with indeter- minate annual fecundity. Results from the growth model indicated that growth rates in males and females were similar. However, the asymptotic length (LOT) predicted for females (1086 mm) was higher than that for males (913 mm). Of Fish sampled, about 75% were greater than 600 mm and over 65% were age 5 or older. These results are likely a product of the narrow size range (c. 650-1013 mm) of Fish captured by the commercial gill net Fishery, which targets adult 0.16 | 0.14 O 0.12 I «., | 0.08 | 0.06 ^ 0.04 T3 a 0.02 oa 0 1-2, 2-3 2-3, 3-4 3-4, 4-5 4-5, 5-6 5-6, 6-7 6-7, 7-8 Width Differences Assessed Fig. 7. Differences in adjacent interannular widths (± 1 SE; n = 527, df = 6, 520, F-ratio = 86.691, p < 0.001) in Gulf corvina; growth rate decreased at the end of the second year, indicating the onset of maturity. 12 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Fig. 8. Mean diameter (± SE) of the four developmental stages of Gulf corvina oocytes; data labels indicate sample sizes. Due to similarities in oocyte diameter, Stage 3 and 4 oocytes were combined to estimate batch fecundity. corvina during their spawning season. Alternatively, it may be that females do reach a larger maximum length, which has been seen in congeners (e.g., C. nebulosus ; Ihde 2000). The overall growth pattern of Gulf corvina is characterized by rapid early life growth that slows a couple years after the onset of sexual maturity. A reduction in population size is commonly accompanied by a phenotypic response of earlier maturation and faster growth rate (Hutchings 2005; Policansky 1993). Associated benefits of early maturation include increased likelihood of reproducing before harvest and greater fitness (i.e., more chances to spawn; Hutchings and Baum 2005). However, earlier maturity may also result in reduced fecundity, decreased post-spawning survival, potentially less fit offspring, and increased probability of negative population growth (Hutchings and Baum 2005; Berkeley et al. 2004). The large size that Gulf corvina reach in such a short time is similar to other sciaenids that co-occur in the Gulf of California ( Cynoscion xanthulus, Riedel et al. 2007; Totoaba macdonaldi, Roman-Rodriguez and Hammann 1997), with rapid growth early in life, reaching an average length of 279 mm by age 2 and 359 mm by age 3 (Figure 5). This is likely a result of the high productivity that characterizes the northern Gulf of California (Brusca 2010), since fish production is largely dependent on primary production (Ryther 1969; Dailey et al. 1993) and food availability tends to be positively correlated to increased growth rates of fishes (Weatherley 1972). The Gulf has extraordinarily high levels of primary productivity, due mainly to upwelling, tidal mixing, and thermohaline circulation ( Alvarez-Borrego 2010; Zeitzschel 1969), which in-turn supports some of the world’s largest sardine, anchovy, and shrimp populations and fisheries (Alvarez-Borrego 2010), all of which are food for Gulf corvina (Roman-Rodriguez 2000). Similarly, positive correlations between growth rate and water temperature are common in sciaenids and many other marine fishes (Lankford and Targett 1994; Brander 1995; Williams et al. 2007), and sea surface temperatures in the upper Gulf of California average 24° C and exceed 32° C during the summer months. Our initial estimates of the growth parameter (k) of 0.255/yr appeared relatively low for a species with a fast growth rate and a maximum age of 8 yrs, although similar growth rates have been reported in populations of Cynoscion species (Shepherd and Grimes 1983; Riedel et al. 2007; Colura et al. 1994). In addition, we found clear differences in the GROWTH AND REPRODUCTION OF C OTHONOPTERUS 13 2000 2500 3000 3500 4000 4500 Fig. 9. (left) Ordinary least-squares regressions of batch fecundity and total length for 2009, 2010, and grouped for Gulf corvina. (right) Ordinary least-squares regressions of batch fecundity and gonad-free body weight for 2009. 2010, and grouped for Gulf corvina. For each graph, the solid black line represents the line of best fit, and the red dashed lines indicate 95% confidence intervals. pattern of residuals between the fishery and bycatch data, which indicated that fish sampled from the targeted corvina fishery tended to be larger at age than those collected as bycatch from other fisheries. These dissimilarities may reflect that these samples were collected from two subpopulations of Gulf corvina with different growth strategies. For example, the targeted Gulf corvina fishery operates in the estuaries of the Colorado River Delta, whereas the shrimp fishery that provided most of our smaller samples operates mainly at offshore locations and habitats 50 km or more to the south or southeast (Rodriguez-Quiroz et al. 2010). Variations in growth rate were reported among populations of the weakfish ( Cynoscion regalis) in the Middle Atlantic Bight and were 14 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES attributed to varying allocations of energy to somatic growth in relation to environmental and migratory requirements and the availability of food items in each habitat (Shephard and Grimes 1983). Alternatively, such differences may simply be a product of the size selectivity of each fishery. Shrimp and croaker fisheries use 6. 5-8. 5 cm mesh gill nets that are deployed as drift nets and tend to land corvina less than 50 cm long. Conversely, the Gulf corvina fishery uses 14.6 cm mesh gill nets deployed in a similar fashion to purse seines (fish are encircled; see Erisman et al. 2012), and selectively removes fishes 50 cm or greater in length. Under this scenario, fish harvested by the corvina fishery may be larger than average size-at-age, whereas those landed as bycatch may be average or below average in size. Regardless of the cause, the two sources of samples included in the age and growth study produced a VBGM with significant uncertainty, and one that may not accurately reflect the average growth rate of the species or incorporate possible variations in growth rate by location. It is important to produce unbiased life history parameters that accurately reflect the stock or population, as these feed directly into fisheries assessments (e.g., estimates of natural mortality and exploitation rates; virtual population models, calculations of sustainable yield). Even subtle changes to these life history parameters may result in significant changes in model outputs and conclusions (Chen et al. 1992; Van Den Avyle and Hayward 1999), which in turn, can result in highly different perceptions among resource managers with respect to the health or condition of the stock and changes in harvest regulations. The systematic collection of representative length samples should be implemented by a scientifically designed, fishery-independent survey that utilizes nets with a variety of mesh sizes during several periods of the year and at several locations. Results from this type of survey would provide a better representation of the overall length and age structure of the population (Van Den Avyle and Hayward 1999). Histological analyses of male and female gonad tissues indicate that Gulf corvina reaches sexual maturity at approximately 2 yrs of age. In addition, differences in adjacent otolith growth increment widths indicate that after age two, otolith growth slows significantly, and all subsequent ages are characterized by similar slow growth rates. Tight correlation between fish size and otolith size (radius, length or weight) suggest that otolith growth rates are an appropriate proxy for estimating fish growth rates; however these relationships are nonlinear and length backcalculations must account for this (Campana and Jones 1992; Pannella 1971; Secor and Dean 1989). Measurements of otolith length, radius or width are often used in fisheries as corollaries to fish size and to back calculate fish age. While in some more fine resolution studies, using these proxies may not be appropriate, as demonstrated by Mosegaard et al. (1988) in 8-108 day experiments; however, in a lower resolution context (years vs. days or months) otolith growth appears to scale with somatic growth sufficiently to make estimates. In this study we find that Gulf corvina otolith size and weight are correlated to fish size, and otolith growth (radius) is highly representative of fish growth curves. The onset of maturity as seen in the slowing of otolith growth rates lagged that of histological analyses by a couple years. This suggests that for fish with life history characteristics similar to Gulf corvina, analyzing otoliths growth rates may provide additional archival information on age at maturity. Fish ranging from 650 to 750 mm and 5 to 6 yrs of age were the most common in the fishery, with few fish larger than 800 mm or 8 yrs of age captured by the gillnet fishery. It is uncertain whether Gulf corvina reach lengths and ages greater than those harvested by the fishery or whether the fishery harvests the largest individuals of this species. Thus, GROWTH AND REPRODUCTION OF C OTHONOPTERUS 15 females are harvested after an average of three years of spawning (age 5 yrs) and at only — 50% of their potential length and therefore, with much lower reproductive output. A larger size enables females to produce more oocytes because fecundity is correlated to body size (e.g., a 1400 mm female Gulf corvina could produce over 2,500,000 eggs per batch). Additionally, older females have been shown to produce higher quality oocytes and larvae that are more resistant to starvation and grow faster than those produced by younger fish (Berkeley et al. 2004). The distribution of oocyte diameters and the presence of several developmental stages within samples indicate that Gulf corvina is a multiple batch spawner (Calliet et al. 1996; Macchi 1998). Multiple stages were present in all samples with 57 of 58 samples consisting mostly (>75%) of oocytes of one developmental stage. The presence of multiple developmental stages within a single ovary indicates that these are indeterminate spawners (Hunter et al. 1985). Also, some females captured in the morning (i.e., before spawning), showed clear histological evidence of daily spawning via the presence of both recent (<12 hr) POF's and newly hydrating oocytes. Therefore, total annual fecundity is not fixed at the beginning of the reproductive season, and immature oocytes continually mature and are spawned throughout the reproductive season (Hunter et al. 1992). In order to estimate total annual fecundity for multiple batch spawning fishes, information on both the number of times an individual spawns in a reproductive season and the percent of oocytes that are resorbed are necessary (Hunter et al. 1985). For this, a tagging study is needed to determine the number of times a female returns to the estuary to spawn in a single year. Batch fecundity estimates ranged from 240,394 to 1,219,342 eggs with a mean of 684,293 eggs per spawn. Roman-Rodriguez (2000) estimated a similar BF range of 250,000 to 808,000 eggs per batch for C. othonopterus. However, only nine fish were sampled in the previous study, which may have resulted in a narrower range. Our fecundity estimates differed significantly between years, with 2010 having a significantly greater fecundity. This could be due to the El Nino event that occurred from May 2009 through April 2010. During and after El Nino events significant increases in zooplankton biomass have been documented in the Gulf of California (Jimenez-Perez and Lara-Lara 1988; Lavaniegos-Espejo and Lara-Lara 1990; Sanchez-Velaso et al. 2000). Augmented zooplankton biomass would increase the productivity of the area and likely result in a greater abundance of food for C. othonopterus , which would provide more energy for reproduction. Fifty percent of females mature at just less than 300 mm but the fishery targets fish greater than 600 mm. Therefore, a complete range of mature females was not used to estimate BF and fecundity is not known for smaller mature females. Fishery- independent sampling of smaller samples is needed in order to understand the correlation between BF and length at the onset of maturity. Acknowledgements We would like to acknowledge Jose Campoy, an amazing advocate for the upper Gulf of California, rest in peace. We thank Martha Roman, an expert on the biology and fishery of the Gulf corvina, for her insights. T. Gedamke and A. MacCall contributed insights and analyses related to the growth models. Thanks to G Paredes, 1 Mascarenas- Osorio, M. Calderon, J Cota-Nieto, and S Sanchez for their assistance with fieldwork and M. Franklin and M. Steele for their thoughtful comments on this research. We especially thank the fishers of El Golfo de Santa Clara and El Zanjon and Alto Golfo Sustentable (AGS) for their support and commitment to this project. The Walton Family 16 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Foundation, the David and Lucile Packard Foundation, the Nearshore Marine Fish Research Program at California State University Northridge, NOS- Noroeste Sustentable A.C., and the Center for Marine Biodiversity and Conservation (CMBC) at Scripps Institution of Oceanography, University of California San Diego provided financial or other support for this research. Literature Cited Allen, L.G., T.E. Hovey, M.S. Love, and J.T.W. Smith. 1995. The life history of the spotted sand bass ( Paralabrax maculatofasciatus) within the southern California bight. California Cooperative Oceanic Fisheries Investigations Reports, 36:1 11. Alvarez-Borrego, S. 2010. Physical, chemical, and biological oceanography of the Gulf of California. Chapter 2 in The Gulf of California, biodiversity and conservation. (R.C. Brusca, ed.) University of Arizona Press, Tucson, Arizona. Pp. 24M8. Berkeley, S.A., M.A. Hixon, R.J. Larson, and M.S. Love. 2004. Fisheries sustainability via protection of age structure and spatial distribution offish populations. Fisheries, 29(8): 23-32. Brander. K..M. 1995. The effect of temperature on growth of Atlantic cod (Gadus morhua L.). Journal of Marine Science, 52:1-10. Brown-Peterson, N.. S. Lowerre-Barbieri. B. Macewicz, F. Saborido-Rey, J. Tomkiewicz, and D. Wyanski. 2007. An improved and simplified terminology for reproductive classification in fishes. Gonadal Histology Workshop of the Joint Meeting of Ichthyologists and Herpetologists; July 2007: St. Louis, Missouri. Brusca, R.C. 2010. Introduction in The Gulf of California, biodiversity and conservation. (R.C. Brusca, ed.) University of Arizona Press, Tucson. Arizona. Pp. 1-5. Cailliet, G.M.. M.S. Love, and A.W. Ebeling. 1996. Fishes: A field and laboratory manual on their structure, identification, and natural history. Belmont: Wadsworth Publishing Company. 194 pp. Campana. S.E. and C.M. Jones. 1992. Analysis of otolith microstructure data in Otolith microstructure examination and analysis. (D.K. Stevenson and S.E. Campana, eds.) Canadian Special Publication of Fisheries and Aquatic Sciences 1 17. Pp. 73-100. Campoy, J.R. and M.J. Roman-Rodriguez. 2002. Observaciones sobre la biologia y pesqueria de la curvina golfina (Cynoscion othonopterus) en la reserva de la Biosfera Alto Golfo de California y Delta del Rio Colorado. Documento de Trabajo, INE IMADES. 1-13. Chao, N.L. 2003. Taxonomy of the seatrout, genus Cynoscion (Pisces, Sciaenidae), with artificial keys to the species. Sciaenidae, in Biology of the Spotted Seatrout. (S.A. Bortone, ed.) CRC Press, Boca Raton, FL. Pp. 5-15. Charnov, E.L., T.F. Turner, and K.O. Winemiller. 2001. Reproductive constraints and the evolution of life histories with indeterminate growth. Proceedings of the National Academy of Sciences, 98:9460-9464. Chen, Y., D.A. Jackson, and H.H. Harvey. 1992. A comparison of von Bertalanffy and polynomial functions in modeling fish growth data. Canadian Journal of Fisheries and Aquatic Sciences, 49: 1228-1235. Colura, R.L., A. Mach, and R. Buckley. 1994. Comparison of Texas spotted seatrout growth among year classes and bay systems, Final Report for Federal Aid in Fisheries Restoration Act Project F-36-R, in Marine Fisheries Resources Culture and Enhancement, Texas Park and Wildlife Department, Austin, TX. CONAPESCA. 2010. Subdelegacion de Pesca en Sonora. En: Manual de capacitacion y diseno de recursos sobre diferentes tecnicas recomendadas para el manejo y procesamiento de la curvina golfina que permitira dar el valor agregado requerido, que sera diverso y muy relacionado a los habitos de consumo de los diferentes mercados en los que se pretende coiocar el recurso (producto 3). Consultores Ejecutivos de Sonora S.A.de C.V. Agosto de 2010, Guaymas, Sonora 2010. 84 pp. Craig, M.T., D.J. Pondella, II., and J.C. Hafner. 1999. Analysis of age and growth in two eastern Pacific groupers (Serranidae: Epinephelinae). Bulletin of Marine Science, 65(3): 807-814. Dailey, M.D., D.J. Reish, and J.W. Anderson. 1993. Ecology of the Southern California Bight: A synthesis and interpretation. University of California Press, Berkeley, California. 926 pp. DOF [Diario Oficial de la Federation], 2005. ACUERDO por el que se establece veda temporal para la captura de curvina golfina (Cynoscion othonopterus) en las aguas marinas y estuarinas de jurisdiction federal de la reserva de la Biosfera Alto Golfo de California y Delta del Rio Colorado, durante el periodo del 1 de mayo al 3 1 de agosto de cada ano. Pub. DOF. 25 de Agosto de 2005, 2 pp. GROWTH AND REPRODUCTION OF C. OTHONOPTERUS 17 DOF [Diario Oficial de la Federation]. 2007. NORMA Oficial Mexicana, NOM-063-PESC-2005, pesca responsable de curvina golfina ( Cynoscion othonopterus) en aguas de jurisdiction federal del Alto Golfo de California y Delta del Rio Colorado. Especificaciones para su aprovechanriento. Pub. DOF, 16 de Agosto de 2007, 7 pp. Erisman, B.E., I. Mascarenas, G. Paredes, Y. Sadovy de Mitcheson, O. Aburto-Oropeza, and P.A. Hastings. 2010a. Seasonal, annual, and long-term trends in commercial fisheries for aggregating reef fishes in the Gulf of California, Mexico. Fisheries Research, 106:279-288. — , M.T. Craig, and P.A. Hastings. 2010b. Reproductive biology of the Panama Graysby, Cephalopliolis panamensis (Teleostei: Epinephelidae), from the Gulf of California. Journal of Fish Biology, 76:1312-1328. O. Aburto-Oropeza, C. Gonzalez-Abraham, I. Mascarenas-Osorio, M. Moreno-Baez, and P.A. Hastings. 2012. Spatio-temporal dynamics of a fish spawning aggregation and its fishery in the Gulf of California. Scientific Reports 2: 284 / D01:10.1038/srep00284 Humason, G.L. 1972. Animal tissue techniques. 3rd ed. W.H. Freeman and Company Publishers, San Francisco, California. 641 pp. Hunter, J.R., N.C.H. Lo. and R.J.H. Leong. 1985. Batch fecundity in multiple spawning fishes in An egg production method for estimating spawning biomass of pelagic fish: application to the northern anchovy (Engraidis mordax). (R. Lasker, ed.) U.S. Department of Commerce, National Oceanic and Atmospheric Administration Technical Report National Marine Fisheries Service, 36:67-78. — and B.J. Macewicz. 1985. Rates of atresia in the ovary of captive and wild northern anchovy, Engraidis mordax. Fishery Bulletin, 83:1 19-136. — , , and C.A. Kimbrell. 1989. Fecundity and other aspects of the reproduction of sablefish, Anoplopoma fimbria , in central California waters. California Cooperative Oceanic Fisheries Investigations Reports, 30:61-72. — , , N.C. Lo, and C.A. Kimbrell. 1992. Fecundity, spawning, and maturity of female Dover sole Microstomas pacificus , with an evaluation of assumptions and precision. Fishery Bulletin, 90: 101-128. Hutchings, J.A. 2005. Consequences of life-history change to population recovery in Atlantic cod, Gadus morhua. Canadian Journal of Fisheries and Aquatic Sciences, 62:824—832. — and J.K. Baum. 2005. Measuring marine fish biodiversity: temporal changes in abundance, life history and demography. Philosophical Transactions of the Royal Society B, 360:315-338. Ihde, T.F. 2000. Biology of the spotted seatrout in the Chesapeake Bay region. M.S. thesis, Viginia Institute of Marine Sciences, Glouchester Point, Virginia. 121 pp. Jimenez-Perez, J.C. and J.R. Lara-Lara. 1988. Zooplankton biomass and copepod community structure in the Gulf of California during the 1982-1983 El Nino event. California Cooperative Oceanic Fisheries Investigations Reports, 29:122-128. King, J.R. and G.A. McFarlane. 2003. Marine fish life history strategies: applications to fishery management. Fisheries Management and Ecology. 10:249-264. Lankford, T.E. and T.E. Targett. 1994. Suitability of estuarine nursery zones for juvenile weakfish (Cynoscion regalis ): effects of temperature and salinity on feeding, growth and survival. Marine Biology. 119:611-620. Lavaniegos-Espejo, B. and J.R. Lara-Lara. 1990. Zooplankton of the Gulf of California after the 1 982 1983 El Nino event: biomass distribution and abundance. Pacific Science, 44:297-310. Macchi, G. 1998. Preliminary estimates of spawning frequency and batch fecundity of striped weakfish, Cynoscion striatus, in coastal waters off Buenos Aires province. Fishery Bulletin, 96:375-381. Macewicz, B.J. and J.R. Hunter. 1994. Fecundity of sablefish, Anoplopoma fimbria, from Oregon coastal waters. California Cooperative Oceanic Fisheries Investigations Reports, 35:160-174. Mosegaard, H., H. Svedang, and K. Taberman. 1988. Uncoupling of somatic and otolith growth rates in Arctic char (Salvelinus alpinus) as an effect of differences in temperature response. Canadian Journal of Fisheries and Aquatic Sciences, 45:1514—1524. Musick, J.A.. M.M. Harbin, A. Berkeley, G.H. Burgess, A.M. Eklund, L.T. Findley, R.G. Gilmore, J.T. Golden. D.S. Ha, G.R. Huntsman, J.C. McGovern, S.J. Parker, S.G. Poss, E. Sala, T.W. Schmidt, G.R. Sedberry, H. Weeks, and S.G. Wright. 2000. Marine, estuarine and diadromous fish stocks at risk of extinction in North America (exclusive of Pacific salmonids). Fisheries, 25:6-30. Pannella, G. 1971. Fish otoliths: daily growth layers and periodic patterns. Science, 173:1124-1127. Paredes, G.A., B.E. Erisman, O. Mascarenas, J. Cota Nieto, K. Gherard, and O. Aburto Oropeza. 2010. La curvina golfina: biologia, pesqueria y su gente. CONABIO. Biodiversitas, 91:1-5. 18 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Policansky, D. 1993. Fishing as a cause of evolution in fishes in The exploitation of evolving resources. (T.K. Stokes, J.M. McGlade, and R. Law, eds.) Springer-Verlag, Berlin, Germany. Pp. 3-18. Rasband. W.S. 1997-2009. ImageJ. Bethesda: U. S. National Institutes of Health. Available: http://rsb. info.nih.gov/ij/ Ratkowsky, D.A. 1983. Nonlinear regression modeling. Marcel Dekker Incorporated, New York. 276 pp. Riedel, R., L.M. Caskey, and S.H. Hurlbert. 2007. Length-weight relations and growth rates of dominant fishes of the Salton Sea: implications for predation by fish-eating birds. Lake and Reservoir Management, 23:528-535. Robertson, D.R. and G.R. Allen. 2008. Shorefishes of the tropical eastern Pacific online information system. Balboa, Panama: Smithsonian Tropical Research Institute. Available: http://biogeodb.stri. si.edu/sftep/. Version 1.0. Roman-Rodriguez. M.J. 2000. Estudio pobacional del chano norteno, Micropogonias megalops y la curvina Golfina Cynoscion othonopterus (Gilbert) (Pisces: Sciaenidae), especies endemicas del alto Golfo California, Mexico. Instituto del Medio Ambiente y Desarrollo Sustentable del Estado de Sonora. Informe final SNIB-CONABIO proyecto No. L298. Mexico D.F. 143 pp. — and M.G. Hammann. 1997. Age and growth of Totoaba, Totoaba macdonaldi (Sciaenidae), in the upper Gulf of California. Fishery Bulletin, 95:620-628. Rowell, K., K.W. Flessa, D.L. Dettman, and M. Roman. 2005. The importance of Colorado River flow to nursery habitats of the Gulf Corvina. Canadian Journal of Fisheries and Aquatic Sciences, 62: 2874-2885. — , — , — — , M.J. Roman, L.R. Gerber, and L.T. Findley. 2008. Diverting the Colorado River leads to a dramatic life history shift in an endangered marine fish. Biological Conservation, 141: 1138-1148. Rodriguez-Quiroz, G., E.A. Aragon-Noriega, W. Valenzuelu-Quihonez, and H.M. Esparza-Leal. 2010. Artisanal fisheries in the conservation zones of the Upper Gulf of California. Revista de Biologia Marina y Oceanografia, 45(1): 89-98. Ryther, J.H. 1969. Photosynthesis and fish production in the sea. Science, 166:72-76. Sadovy, Y. and D.Y. Shapiro. 1987. Criteria for the diagnosis of hermaphroditism in fishes. Copeia, 1:136-156. Sadovy de Mitcheson, Y. and M. Liu. 2008. Functional hermaphroditism in teleosts. Fish and Fisheries, 9:1-43. Sanchez-Velaso, L., B. Shirasago, M.A. Cisneros-Mata, and C. Avalos-Garcia. 2000. Spacial distribution of small pelagic fish larvae in the Gulf of California and its relation to the El Nino 1997-1998. Journal of Plankton Research, 22(8): 1611 1618. Secor, D.H. and J.M. Dean. 1989. Somatic growth effects on the otolith-fish size relationship in young pond-reared striped bass, Morone saxatilis. Canadian Journal of fisheries and Aquatic Sciences, 46(1): 113-121. Shepherd, G. and C.B. Grimes. 1983. Geographic and historic variations in growth of weakftsh, Cynoscion regalis, in the Middle Atlantic Bight. Fishery Bulletin, 81(4): 803-813. Shin, Y., M. Rochet, S. Jennings, J.G. Field, and H. Gislason. 2005. Using size-based indicators to evaluate the ecosystem effects of fishing. ICES Journal of Marine Science, 62:384-396. Van Den Avyle, M.J. and R.S. Hayward. 1999. Dynamics of exploited fish population, in Inland fisheries management in North America. (C.C. Kohler and W.A. Hubert, eds.) American Fisheries Society, Bethesda, Maryland. Pp. 127-166. Weatherley, A.H. 1972. Growth and ecology of fish populations. Academic Press Incorporated, London, England. 293 pp. Williams, J.P., L.G. Allen. M.A. Steele, and D.J. Pondella, II. 2007. El Nino periods increase growth of juvenile white seabass ( Atractoscion nobilis) in the Southern California Bight. Marine Biology, 152: 193-200. Zeitzschel, B. 1969. Primary productivity in the Gulf of California. Marine Biology, 3:201-207. Bull. Southern California Acad. Sci. 112(1), 2013, pp. 19—37 © Southern California Academy of Sciences, 2013 Recent Mass Mortality of Strongylocentrotiis purpuratus (Echinodermata: Eehinoidea) at Malibu and a Review of Purple Sea Urchin Kills Elsewhere in California Gordon Hendler Natural History Museum of Los Angeles County, 900 Exposition Boulevard, Los Angeles, CA, 90007, USA, hendler@nlun.org Abstract. — Mass mortalities of intertidal purple sea urchins, Strongylocentrotiis purpuratus occurred at Malibu Lagoon State Beach, California, in 2010 and 2011. Both events followed the first heavy rain of the season, and coincided with the illegal breaching of a coastal lagoon. Osmotic shock from low-salinity lagoon water, the likely cause of death, may have acted jointly with stress from subaerial exposure during especially low tides. Massive die-offs of purple sea urchins have occurred at other localities, usually after natural conditions created lethal levels of osmotic or thermal stress, or because of human efforts to harvest or to eradicate the species. Annually recurring lagoon ruptures at Malibu, combined with predation by western gulls, can have a profound impact on the local population of S. purpuratus and on intertidal ecology. Introduction At sunrise on October 9, 2010, in Malibu, California, dignitaries inaugurating Surfrider Beach (Malibu Lagoon State Beach) as the first World Surfing Reserve unexpectedly encountered more than 10,000 freshly killed Strongylocentrotiis purpuratus (Stimpson, 1857) stranded at the tideline (Ralston, pers. comm.). The cause of their death was still undetermined a year later, after another, less severe mass mortality of sea urchins occurred at the same locality. What killed the sea urchins at Malibu in these two successive years, and what might cause mass mortalities of S. purpuratus elsewhere in California? Strongylocentrotiis purpuratus has a broad range of tolerance for thermal, osmotic, and anaerobic stress as evidenced by its extensive distribution from Shelikof Island. Alaska to Isla Cedros, Mexico, and from the mid-intertidal to 30 m (exceptionally to 161 m) depths (Sagarin and Gaines, 2002; Lambert and Austin, 2007; Lester et al., 2007; Pearse and Mooi, 2007; Ebert, 2010, and pers. comm.). Yet its habitat sometimes imposes stresses that exceed the species’ tolerance, resulting in die-offs of purple sea urchins. Past events of this sort generally have been attributed to radical alterations in salinity or temperature, or disease. Since the same factors could have been responsible for the mass mortality at Malibu Lagoon, I discuss the specific tolerances and vulnerabilities of S. purpuratus , and review the causes of prior massive kills. Based on that information, I offer a hypothesis regarding the events at Malibu. Individuals of S. purpuratus are weakened or killed by exposure to extreme temper- atures or salinities (Schroeter, 1978). They cannot survive for more than 3 h in seawater that is < 60% ( — 20%o) or > 120% (~39%o) in concentration, and they die after 4 h, if exposed to temperatures more extreme than 1.9 to 23.5 C (Farmanfarmaian and Giese, 1963; Giese and Farmanfarmaian, 1963; Burnett et al., 2002); vulnerability to elevated temperatures may shape the southern range limit of S. purpuratus (Ebert, 2010). Fike 19 20 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES other echinoids that have been tested, purple sea urchins probably can tolerate, for hours or days, mild to moderate hypoxia (> 0.5 ml DO L ’ 1 ) or elevated levels of hydrogen sulfide, but cannot survive severe hypoxia or anoxia (< 0.5 ml DO L_l) (Thompson et al., 1991: Riedl et ah, 2012). They withstand hypoxic conditions by curbing their oxygen consumption and exploiting anaerobic respiration, and individuals that are emersed can absorb oxygen from the air (Johansen and Vadas, 1967; Burnett et al., 2002). Although S. purpuratus tenaciously grips the substrate to avoid dislodgement (Denny et al., 1985; Denny and Gaylord, 1996), large numbers of individuals have been torn from rocky reefs by wave surge under particularly harsh winter-storm conditions. Abrupt changes in salinity during storms have also proved lethal to A. purpuratus. Ebeling et al. (1985) reported that an extensive “urchin front” off Santa Barbara, California, was decimated by surge during a violent storm in 1983. A severe storm in 1938 ruptured a river levee, releasing a massive volume of fresh water into Newport Bay, California, killing animals as far as 3.2 km from the mouth of the bay (MacGinitie, 1939:685). Purple sea urchins were “among those animals which were most obviously devastated by the storm... hundreds of the tests were washed up on the beach or fell to the bottom between the rocks where they were visible at low tides as whitish patches. Even at the lowest tide not a single sea urchin could be seen on the rocks.” Despite the massive fatalities, the population nearly recovered within a year. A flash flood in 1977, which deluged the rocky shore at Corona del Mar, California, killed 90.5% of S. purpuratus (a decrease of 60 per nr) in the mid- and lower intertidal of the affected area (Littler and Littler, 1987). Windrows of urchins were found cast onto the beach two days after the storm, at which time salinity measured a record low 23.5%o. Sousa (1979) found a reduced density of S. purpuratus and crabs in an intertidal boulder field near Santa Barbara, California, where freshwater runoff from a seasonal creek reached the intertidal. In 1973, at the same locality, Schroeter (pers. comm.) observed that “a single large rainstorm caused freshwater discharge that ponded in the lower portion of the boulder field creating a purple sea urchin bouillabaisse sans fish and near complete mortality. All of the urchins that were at a slightly higher elevation (—0.1- 0.2 feet higher) appeared to be unaffected.” A similar rain related die-off was thought to have occurred at Sunset Bay, Oregon, where slow recovery of the population of S. purpuratus was anticipated (Ebert and Grupe, 2008). Thermal stress has also caused its share of mass mortalities of S. purpuratus. Hedgpeth and Gonor (1969:94) recorded internal body temperatures of S. purpuratus “...of 26° C...in beds of urchins exposed to the sun, and maximums of 27° C. to 30 C. in some urchins. Successive days of this type of heating at low tide for periods of three to five hours led to a heat kill at the study site and other areas along the central Oregon coast, with many urchins dying in each place.” They also noted that similar die-offs of S. purpuratus had been observed at other localities in Oregon for years. A mass mortality event coincident with water temperatures > 24 °C decimated S. purpuratus at False Point, La Jolla, California, in the summer of 1971. Sea urchins did not reappear there until the winter of 1973, following the successful settlement of juveniles (Ebert, 1983). In another instance, several hundred dead S. purpuratus were found after a series of minus tides at Newport, California, and the mortality event was tentatively attributed to 24 to 27 °C water temperatures (So, 2006).* *So, H. 2006. Officials suspect weather played a role in mass sea urchin deaths. Los Angeles Times. August 3, 2006:B.3. Available from http://articles.latimes.com/2006/aug/03/local/me-seaurchin3 via the Internet. Accessed 2 November, 2012. PURPLE SEA URCHIN MASS MORTALITY 21 In addition, disease outbreaks have been reported among 5. purpuratus in California, and increasingly frequently among other echinoid species, during the last several decades (Ward and Lafferty, 2004). In the 1970s, infected Strongylocentrotus franciscanus (A. Agassiz, 1863) were first found near San Diego, and afterwards at localities off central California (Johnson, 1971; Pearse et al., 1977). Affected animals could not grip the substrate, their ossicles deteriorated and spines detached, and they died within several months. However, the affected populations recovered (Pearse et ah, 1977; Pearse and Hines, 1987). In the same region, Gilles and Pearse (1986) found a low incidence of diseased S. purpuratus. Lester et ah (2007) documented two different pathologies of infected S. purpuratus , presumed to represent distinct diseases, in an intertidal survey spanning central California to Baja California, Mexico in 2004. They found that disease prevalence was positively correlated with temperature but not with population density, leading them to suggest that “If thermal stress does increase host susceptibility, disease in Strongylocentrotus species could increase with global warming and increased human impacts along the coast” (Lester et ah, 2007:322). Yet another source of mortality, thus far manifested by a single, massive die-off of S. purpuratus and less drastic damage to S’. franciscanus , appears to have been produced by an unidentified toxin from a harmful algal bloom in northern California during 201 1 (Jurgens, pers. comm.). With respect to the mass mortalities at Malibu State Beach. California, in 2010 and 2011, I propose for reasons discussed herein that the sea urchins were killed by fresh water released through a breach in a rain-swollen coastal lagoon, which drastically lowered salinity in the intertidal zone. I also suggest that damage from osmotic stress was probably aggravated by subaerial exposure of the animals during low tides, and possibly by their burial under transported sediment. Materials and Methods Study Site At present, Malibu Lagoon is an ecologically dysfunctional seasonal estuary. Roughly 0. 1 km2 in area, it channels water from a surrounding 284 km2 drainage basin into Santa Monica Bay (Figure 1A). The lagoon is the remnant of a previously much larger estuary and wetland that, after 1900, was drastically reduced in size by repurposing natural habitat for ranching and urban development, and by constructing a dam, sewage treatment plant, railroad, and highway, all of which disrupted the estuarine lagoon system and the ecology of its biota (Ambrose and Orme, 2000). As a result “...the Malibu estuarine lagoon is no longer a natural system because, although stream floods and storm waves may sometimes reassert dominance there are now so many constraints imposed by human activity” (Ambrose and Orme, 2000:2-4; Schwarz and Orme, 2005). The lagoon changes seasonally from closed lagoon to estuary. Typically during the summer and fall, drifting sediment forms a sand bar that blocks the estuary’s mouth, allowing fresh water from the City of Malibu and Malibu Creek to accumulate in the embayment (Ambrose and Orme, 2000). Floods from winter storms rupture the sand bar, flushing brackish water, sediment, biota, nutrients, and human and animal waste from the lagoon through the breach (Ambrose and Orme, 2000; Schwarz and Orme, 2005). Usually throughout the spring, an opening in the sand bar is maintained by outflow from Malibu Creek, and by tidal flow that can move approximately 100,000 nr of water through the mouth in single tidal cycles (Ambrose and Orme, 2000). The volume may double during extreme weather conditions when the lagoon fills to the highest level (Ambrose and Orme, 2000; Schwarz and Orme, 2005). Seaward of the sand bar there is a SOUTHERN CALIFORNIA ACADEMY OF SCIENCES 22 Fig. 1 . Malibu Lagoon State Park and its contiguous intertidal reefs. A) Aerial view of Malibu Lagoon at high tide on September 5, 2008. Arrows indicate approximate positions of breaches through the sandbar in 2010 (a), and 201 1 (b). B) Portion of Basin Reef at low tide on October 27, 201 1, showing water pooled in the reef s shallow central depression, and Santa Monica Bay in the distance. C) Bird Reef in its entirety at low tide on October 15, 2011. Sediment near the cobbles was transported and redistributed after the breach on October 6, 201 1. D) View of Bird Reef on October 27, 201 1, showing remnants of a channel through the sandbar, which was excavated by outflow through the breach. (Photo A, courtesy H. Burdick, Los Angeles Waterkeeper). wave-washed shelf composed of cobbles and sediment, part of which is exposed subaerially at low tides. Algal cover on the shelf is quite sparse compared to the dense growth of kelp and other macroalgae in the subtidal zone. Beginning in 1984, the California Department of Parks and Recreation (DPR) mechanically breached the lagoon when water levels were high, in order to avert septic system failures in the adjacent Malibu Colony and Civic Center (Ambrose and Orme, 2000). The agency deliberately opened the sand bar near the western side of the lagoon in an effort to facilitate pedestrian access to the beach and to avoid spoiling the surf break. Breaching the bar at the lower semidiurnal tide maximized the grade between lagoon and ocean, maximized the amount of sediment flushed out, and increased the period that the lagoon remained open. DPR suspended the practice in 1997, when it was deemed a potential threat to tidewater gobies ( Eucyclogobius newberryi Girard, 1854) that had been reintroduced in the lagoon (Swift et al., 1993; Ambrose and Orme, 2000). Since that year, despite a legal prohibition, the western end of the sandbar has repeatedly been breached, usually but not always when a natural breach was imminent (Pfeifer, pers. comm.), by individuals “concerned with lagoon eutrophication and the quality of the surfing en- vironment” (Schwarz and Orme, 2005:97). Methods Supratidal and intertidal zones at Malibu Lagoon State Beach were surveyed to investigate the origin of sea urchin kills and to gauge the impact of lagoon breaches. Information was gathered opportunistically, since mortality events were unanticipated. PURPLE SEA URCHIN MASS MORTALITY 23 By necessity, features of the biota and habitat preceding the events were determined based on the observations of local observers and from recordings made by automated monitoring equipment. After the events on 2010 and 201 1, the effects and magnitudes of the kills were assessed by counting the numbers of freshly dead sea urchins deposited in beach wrack, and by measuring their body size. In addition, the population density and size of surviving animals were quantified by counting and measuring living sea urchins along intertidal transects. Strongylocentrotus purpuratus was surveyed at two adjacent “reefs,” i.e., platforms of cobbles and sediment on the intertidal shelf, which were separated by a channel. Basin Reef encloses a central, sandy depression and is seaward of the middle of the lagoon (Figure IB). Bird Reef is near the western side of the lagoon and to the east of another reef on Malibu Point (Figures 1C, D). Ruptures in the sandbar closing the lagoon directed lagoon water toward Basin Reef in 2010, and toward Bird Reef in 2011 (Figure 1A). After the kill in 2010, I visited Malibu Beach on two different days to examine dead animals that had accumulated in beach wrack and to assess their numbers: October 14 near slack tide, one week after the first mass mortality, and during low tide on October 23 (0 cm tide; referenced to Mean Lower Low Water (MLLW)). I also laid transect lines to measure the density of surviving sea urchins at Basin Reef and Bird Reef on November 5 and 7 (during —30 cm tides). At each 15 m long transect line, I counted animals that were exposed, or found under easily turned rocks, in fifteen 0.25 nr quadrats spaced at 1 m intervals. One week after the kill in 201 1, 1 examined the shore, looking for dead animals during low tide on October 13 (+9 cm tide), and once more on 15 October (+18 cm). I also surveyed the number of sea urchins in transects at Basin and Bird reefs on October 26, 27, and 28 (-30, -37, and —34 cm). Additionally, on October 27, 201 1, 1 recorded the size of S. purpuratus at Basin Reef and Bird Reef. Calipers were used to measure the test diameter of sea urchins that survived the event, which were collected from fifteen 0.25 itt quadrats at one of the multiple transects on each reef. On October 13, 2011, I similarly measured the test diameters of urchins that died during the event, which were collected arbitrarily from wrack on the high water line, in order to compare their size to that of live animals in the intertidal zone. Stranded sea urchins were also examined for signs of physical damage in order to determine the cause of their death. Diseased S. purpuratus have distinctive lesions in the integument, discoloration of the test, or “bald” patches that are bare of spines (Lester et al., 2007). Lobsters consume purple sea urchins in their entirety, sea stars denude them of spines, and fishes damage them in a characteristic manner (Tegner and Dayton, 1981). Avian predators break open the dorsal or ventral surface of the test, or rupture the peristomial membrane and remove Aristotle’s lantern (Hendler, 1977; this report). In contrast, individuals exposed to lethal levels of osmotic stress initially become immobile, and remain virtually intact after they die (Giese and Farmanfarmaian, 1963). In 2010, records of water level and water quality in Malibu Lagoon were used to infer the timing of the breach and properties of the water that was released. Water elevation relative to NAVD88 and water quality characteristics were recorded at several stations in the lagoon by Heal the Bay using YSI600XLM sondes, attached 6 to 15 cm above the benthos, which logged data at 30 min intervals (2NDNATURE, 2010). Data from the probes that were functional during 2010 were provided by M. Abramson, Santa Monica Bay Restoration Foundation. 24 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Weather and sea state are not monitored at Malibu State Beach, necessitating the use of proxy data from nearby localities to evaluate the environmental conditions to which purple sea urchins were exposed during mortality events. Weather, wind, and wave information was based on surface reports for Zuma Beach, California, collected four times daily from 0700 to 1600 h (provided by R. Kittell, National Weather Service). Estimates of the timing and height of tides was based on NOAA tide records for Santa Monica, California. Rainfall and weather information for Los Angeles International Airport (KLAX) was downloaded from MesoWest (http://mesowest.utah.edu/index. html) and supplemented with archival records from the Los Angeles Times. Accounts and photographs provided by numerous observers were used to reconstruct the nature and timing of the events at Malibu Beach. The sources are cited in Bindings and Acknowledgements sections, and in figure legends. Results Observations on the October 2010 Mass Mortality At 0600 h on October 12, 2010, the Los Angeles County Department of Beaches and Harbors (DBH) staff encountered “thousands of small purple sea urchins” at Malibu Lagoon State Beach. The beach maintenance crew disposed of some dead animals and raked others into piles. DBH contacted the Natural History Museum of Los Angeles County for information, out of concern regarding the cause of the mortality. When I examined the beach on October 14, dead sea urchins were strewn along the high tide line between the Malibu Lagoon and Malibu Pier. Besides S. purpuratus , there were small numbers of S. franciscanus, approximately 20 Aplysia vaccaria Winkler, 1955, and the remains of several crabs (Loxorhynchus sp.. Cancer sp. ). DBH personnel said they had never before seen a similar event. They mentioned that the lagoon had breached and drained the previous week but, unaware of the time that the sea urchins died, made no connection between the breach and the mortality. Other individuals on site with whom I spoke at the time, and the news media, suggested that the mortality was related to pollution in the lagoon or to a change in water chemistry. Some explanations offered were manifestly incorrect, such as that dead sea urchins had washed out of the lagoon or had been deposited on the beach by seagulls. Initially, I anticipated that the sea urchins may have died from one of the diseases that periodically kill S', purpuratus and S. franciscanus. However, none of the dead animals had the characteristic signs exhibited by diseased strongylocentrotids (Lester et al., 2007). They were desiccated but fairly intact, covered by spines and integument of normal appearance, and many had an unbroken peristomial membrane and Aristotle's lantern. The vast majority were so little damaged that they obviously had not been killed by predators (e.g. Tegner and Dayton, 1981). However, western gulls ( Larus occidentalis Audubon, 1839) on the beach were observed probing and removing tissue from dead sea urchins. To determine the time of the breach and the mortality, I contacted G. Pfeifer, an Ocean Lifeguard Specialist who had worked at the beach for 32 years. His daily log indicated that the lagoon was artificially breached on October 7 at 0930 h, and he recalled that dead sea urchins washed ashore by the morning of October 8. In addition, lie mentioned that nearly every year varying numbers of dead sea urchins stranded on shore after the lagoon was breached, but that the most recent mortality was unusually large. Other sources later corroborated his recollections. PURPLE SEA URCHIN MASS MORTALITY 25 Fig. 2. Breach in the Malibu Lagoon sandbar, which was artificially created on October 7, 2010. A) Water coursing through the berm at 0947 h, shortly after the breach ruptured. B) Aerial view of the southeast portion of Malibu Lagoon on October 7, 2010 at 1315 h, showing a stream of lagoon water running through the breach, across the sandy beach, and onto the intertidal. A field of intertidal cobbles is visible at the lower left. (Photos courtesy H. Burdick, Los Angeles Waterkeeper). It appears that a chain of events resulting in the mass mortality began on October 6, after the first significant rainstorm of the season. Rainfall in Los Angeles was a record for the date, and discharge from Malibu Creek raised the height and lowered the salinity of water in the lagoon. At Los Angeles Airport, 2.31 cm of rain fell intermittently from 0253 to 1253 h. However, it is not known if rain fell during the low tide ( — 3 cm at 1458 h) that day on the sea urchins in Malibu’s subaerially exposed intertidal zone. On October 7, a channel was dug illegally across the berm at 0915 h according to State Parks Ranger T. Hayden, who issued a citation to the individual responsible. The excavation was made near the midpoint of the lagoon where the berm was only about 1.5 m wide (Ogle, pers. comm.; Figure 1A). Although the action coincided with a maximum high tide of +183 cm at 0841 h, which raised the level of seawater at the sandbar, water in the rain swollen lagoon created a considerable hydraulic gradient. Lagoon water rapidly liquefied and ruptured the sandbar, releasing a food of brown, sediment-laden water that swiftly extended seaward from the beach and moved eastward along shore (Ogle and Pfeifer, pers. comm.; Figure 2A). By 1030 h water level in the lagoon dropped 0.6 m (Burdick, pers. comm.), and at 1 130 h a “torrent” from the lagoon still roiled the bay (Thompson, pers. comm.). At 1200 h, water continued to pour though the breach, which had by now grown to 1 to 2 m in width, although a great deal of water still remained in the lagoon (Barboza, pers. comm.). As the tide fell in the early afternoon, water from the lagoon streamed toward a central depression in Basin reef and past a layer of sediment deposited by the outflow, which in some places was >1 m high (Ogle, pers. comm.; Figure 1 B). An aerial photograph, taken at 1315 h when the tide level was +43 cm, showed lagoon water Bowing across the subaerially exposed intertidal (Figure 2B). There were minus tides that afternoon (-15 cm, 1544 h), and for the next 3 days (—21 cm, 1641 h; —18 cm, 1720 h; —12 cm, 1814 h). Thus, immediately after the breach a low tide exposed purple sea urchins to the air and to hyposaline lagoon water that pooled around the individuals situated in depressions on the intertidal shelf. Sea urchins that were immersed in fresh water, or in fresh water mixed with sea water, would have experienced osmotic stress, and the low tides occurring for several days afterwards repeatedly subjected them to heating and desiccation. 26 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Contemporaneously with the breach event. Heal the Bay sondes in the lagoon recorded decreasing water elevations at stations 1 (near the bridge at State Route 1) and 6 (in a western side-branch of the lagoon's main channel), corroborating information provided by observers at the site. Water level began to drop at 0930 h and continued falling until 1900 h, indicating that lagoon water flowed toward the intertidal zone during low tide. Between 0930 and 1830 h the mean and range (in parenthesis) of water quality data recorded at station 6 were: salinity 1.57%o (1.13-2.23), temperature 21.45 °C ( 1 8.32— 24.12), Oxygen Reduction Potential 24.24 mV (—152.50-61.70), and pH 7.99 (7.65-8.53). Based on records at Zuma Beach, inshore seawater temperatures were approximately 17.8-21.1 C, and air temperature during low tide was 17.8 °C. On the afternoon of October 7 and the mornings of October 8 and 9, waves were from the SW and 0.3-0. 6 m in height. On the morning of October 8, less than 24 hours after the breach occurred, large numbers of dead sea urchins were deposited onshore by the rising tide. The following day, considerably more dead animals accumulated on the beach eastward of the lagoon (see Figure 3A, B), and the number of stranded sea urchins was conservatively estimated at 10,000 on a 140 m long segment of the beach (Ogle, Pfeifer, and Ralston, pers. comm.). Over the next several days, prevailing currents moved some dead sea urchins further east toward Malibu Pier, and a small number reached Carbon Beach (Sikich, pers. comm.; pers. obs.; Figure 3C). Western gulls began to feed on dead 5. purpuratus immediately after they were stranded, and a week later they still scavenged desiccated remains of the sea urchins, usually by removing tissue through the peristomium (Ralston, pers. comm.; pers. obs.; Figures 3D, 4A). Observations on the October 201 1 Mass Mortality Two observers who described the mortality event in 2010 as particularly severe, recalled that in previous years dead sea urchins usually washed ashore after the lagoon breached (Pfeifer and Ogle, pers. comm.). However, other individuals familiar with the locality did not recall seeing dead sea urchins at the lagoon after prior breaches, or did not recall seeing significant numbers of dead sea urchins, or remembered only two die- offs of at most 1,000 sea urchins in the previous 10 years. Events in late 201 1 provided an opportunity to assess the divergent viewpoints. The first heavy rainfall of the season on October 5, 2011, which totaled 2.72 cm at Los Angeles Airport, between 0553 and 0753 h, was comparable to the rain accumulation on October 6, 2010. Runoff lowered salinity to 2 to 5%o at various points in the lagoon where it had been 6%o the week before (Krug, pers. comm., based on refractometer readings). After 1600 h on October 6, the lagoon was artificially breached (Pfeifer, pers. comm.). Presuming that breaching was timed to avoid detection by the authorities, it probably was accomplished during or after the +152 cm high tide at 1838 h on October 6, and possibly before the +12 cm low tide at 01 17 h on October 7. Air temperatures during the period ranged from 14.4 to 15.0 °C, seawater temperatures from 17.2 to 17.8 C, and 0.6- 1.2 m high waves from the SW were recorded at Zuma Beach. The breach, located at the western end of the lagoon near Malibu Colony, directed lagoon water and sediment toward the western flank of Bird Reef, which is exposed to incoming waves from Santa Monica Bay (Figure ID). Following the breach, daily low tides exceeded datum until October 25. Dead sea urchins were washed ashore by the morning of October 7 (Pfeifer, pers. comm.), by a rising tide that reached +146 cm at 0746 h; 1 learned of their presence on PURPLE SEA URCHIN MASS MORTALITY 27 Fig. 3. Sea urchins, primarily Strongylocentrotus pur pur ut us, which were killed in a mass mortality event and stranded on the tideline at Malibu Lagoon State Beach. A) Freshly dead animals deposited by the falling tide on October 9, 2010 at 1519 h. B) Close view of the dead sea urchins in A, which are fully spined and have intact peristomial membrane and Aristotle’s lantern. C) Five days after the breach on October 12. 2010, the assemblage of dead animals extends towards Malibu Pier. D) Dead urchins in beach wrack are still largely intact, although spines are missing on some abraded tests on October 14, 2010. (Photos A, B courtesy M. Ralston/AFP; C courtesy D. Murray, Heal the Bay). October 12 (Krug, pers. comm.). On October 13, six days after the breach, I counted 2,000 dead S. purpuratus on the high tide mark between the eastern boundary of Malibu Lagoon and Malibu Pier. Among them were a small number of S. franciscanus, a Pagettia sp., a Panulirus sp. carapace, and a dead cormorant. The mean test diameter of dead S. purpuratus in a sample (n = 103) haphazardly collected from beach wrack that day, was 49.8 ± 6.0 mm (X ± SD). The peristomial membrane and Aristotle’s lantern were intact in all but three specimens, and only 24 specimens were partially denuded of spines. Evidently, the relatively undamaged specimens were victims of the mass mortality event. Elowever, one animal was entirely denuded of spines, and the ventral side of the test of two others was broken open. The latter, broken individuals had probably been attacked by sea gulls (Figure 4), as evidenced by the type of damage that they exhibited. Indeed, western gulls were observed scavenging dead urchins on the beach before, during and after the +0.9 cm low tide at 1652 h on that same day. On October 15, before and after a + 18 cm low tide at 1811 h, 1 watched airborne western gulls repeatedly release S. purpuratus from their bills, dropping them onto rocks. I also witnessed western gulls in the intertidal zone dislodge and open urchins by penetrating just the peristomium, or by breaking into the ventral or dorsal surface of the test with their bill (Figure 4B-D). Their activities were similar during low tides on October 26-28. Although these observations SOUTHERN CAL IFORNIA ACADEMY OF SCIENCES Fig. 4. Examples of predation by western seagulls (Lams occidentalis) on purple sea urchins (Strongylocentrotus purpuratus ) at Malibu Lagoon State Beach. A) L. occidentalis feeding on freshly stranded, dead sea urchins on October 9, 2010. B) Airborne L. occidentalis attempting to drop and break a sea urchin on intertidal cobbles, October 26, 201 1. C) L. occidentalis preying on intertidal S. purpuratus at dusk on October 15, 201 1 . D) Remains of S. purpuratus eaten by a seagull that fractured the dorsal surface of the test, October 15, 2011. (Photo A, courtesy M. Ralston/AFP). were made during the first important series of afternoon minus tides since the previous winter, at a time that numerous S. purpuratus might have been expected to be present, few live purple sea urchins were seen in the mid-intertidal. However, many that were found were recent victims of gull predation (Figure 4A-D). Population Density and Size Distribution On November 7, 2010, a month after the breach, only one living S. purpuratus was detected in 60 quadrats surveyed on Basin Reef, which was situated in the path of water discharged from the lagoon. The density (X ± SD) of S. purpuratus in the quadrats was only 0.02 ± 0.13 per 0.25 m2, and very few additional sea urchins were seen during an examination of the entire reef. Although comparable measurements of the population density at Basin Reef were not made prior to 2010, I was told that local surfers habitually avoided accidental contact with Basin Reef to preclude injuries inflicted by spines of the abundant purple sea urchins living there before the mass mortality (Ogle, pers. comm.). At the same site, after the breach in 2011, 212 animals were found in 113 quadrats surveyed on October 26 and 27, a mean density of 1.88 ± 1.12 sea urchins per 0.25 m2. The statistically significant difference that was found between years, in the density of S. purpuratus , suggests that the population at Basin Reef was nearly exterminated in 2010, and recovered in 2011 (Mann-Whitney U test, 2-tailed, corrected for ties: U = 2,375.5, PURPLE SEA URCHIN MASS MORTALITY 29 Fig. 5. Size frequency distributions of Strongylocentrotus purpuratus collected at Malibu Lagoon State Beach on October 27, 201 1, from Basin Reef (A) and Bird Reef (B). n = 120, P = 0.001 3). In 2010, Bird Reef was not exposed to water released by the breach, and there were 277 S. purpuratus in the 60 quadrats surveyed there on November 5, a density of 4.62 ± 7.72 per 0.25 nr. However, after Bird Reef was Hooded by lagoon water during the breach in 2011, only 73 51. purpuratus were found there in 60 quadrats surveyed on October 27, a density of 1.22 ± 2.33 sea urchins per 0.25 nr. The significant difference in population density between years, and the lesser number of S. purpuratus found in 2011, suggest that mortality related to the breach had drastically reduced the population density of sea urchins at Bird Reef (Mann-Whitney U test, 2-tailed, corrected for ties: U = 5729.5, n = 173, P < 0.0001). On October 27, 2011, individuals of S. purpuratus were collected from transects at Basin Reef and Bird Reef, and they were measured in order to compare the sizes of animals at these two sites, each of which had been affected by a lagoon breach in a different year. At both reefs the size-frequency distributions of purple sea urchins were bimodal (Figure 5), with the smaller size class probably composed of animals between one and two years of age (year-old S. purpuratus are approximately 17 to 33 mm diameter, per Pearse and Pearse, 1975; Rowley, 1990). Purple sea urchins on Basin Reef, where many animals were killed in 2010, were relatively smaller in diameter than those on Bird Reef, where sea urchins were killed by the breach in 2011 (30.70 ± 13.92 mm test diameter, n — 37 versus 38.49 ± 14.23, n = 31 [X± SD] ). The size distribution of animals at Basin Reef could have been shaped by differential size-related mortality in 2010, or by the sizes of sea urchins immigrating to the reef after the die-off. Although adult purple sea urchins are fairly sedentary, they emerge from hiding and move in response to the 30 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES availability of food and to the presence of competitors and predators (Paine and Vadas, 1969; Dayton, 1975; Schroeter, 1979; Ebeling et al., 1985; Tegner and Dayton, 1991). It is possible that the predominant, larger size class of purple sea urchins at Bird Reef, which were survivors of the 2010 and 2011 mortality events, inhibited the immigration of smaller animals, or perhaps more likely that disproportionately more small animals died during the mortality event. Although the relatively large mean size (49.8 mm diameter) of dead purple sea urchins found in beach wrack argues against the latter interpretation, it also suggests a possibility that small individuals were more readily destroyed or that large animals were preferentially transported ashore. Discussion Mass Mortality oj S. purpuratus at Malibu The behavior and physiology of 5. purpuratus were not directly monitored during and immediately after the population was exposed to lagoon discharge. Also, meteorological and oceanographic conditions and the chemistry of water in the lagoon and the bay were not comprehensively monitored at Malibu Lagoon State Beach during mass mortality events in 2010 and 201 1 . As a consequence the cause of the die-offs cannot be determined with certainty, but causative factors can be inferred from the information that is available. It is notable that the only freshly dead marine animals that were found in significant numbers at Malibu were S. purpuratus and S. franciscanus. Other animals deposited among the sea urchins on the beach were so rare that their occurrence seems unrelated to the mass mortality. The small number of dead Aplysia vaccaria seen in 2010 were likely senescent individuals, as this species is an annual (Audesirk, 1979; Angeloni et al. 1999). Sea urchins alone seem to have been affected, suggesting that they were more susceptible than other organisms to the causative agent(s). In addition, so many individuals of S. purpuratus were killed compared to S. franciscanus that it appears the phenomenon was restricted to the intertidal zone where the numbers of S’, purpuratus far exceed those of S. franciscanus , which is primarily a subtidal species in southern California. Circumstances of the mass mortality events in 2010 and 201 1 were consistent with the hypothesis that the sea urchins were killed by brackish water released through breaches in the lagoon. In both instances, dead sea urchins appeared two days after the first major rainfall of the annual rainy season and one day after the lagoon was breached. In both instances dead animals were deposited down-current from the breach. In both instances it appeared that the population density of S. purpuratus was reduced on the reef closest to outflow from the lagoon, and that the population recovered to some extent within a year. Parallels between events indicate that in both cases sea urchins were killed by the same agents. However, differences in the extent of mortality between years suggest that several interacting factors were involved. Based on the literature (reviewed in the Introduction), the physical factors most often associated with mass mortality of S. purpuratus included exposure to elevated temperature > 23.5 °C, or to reduced salinity < 21%o, for several hours. Since the salinity of water released from the lagoon was only I to 5%o in 2010 and 2011, it could well have injured or killed individuals of S. purpuratus , even if its salinity were moderately elevated by an admixture of seawater. Ambient temperature was unlikely to have been injurious, since nearshore air and water temperatures were below lethal levels in both years. Although water temperatures > 23.5 C were recorded in the lagoon for 2 h during PURPLE SEA URCHIN MASS MORTALITY 31 the 2010 breach, water leaving the lagoon would have cooled after exposure to the air and to sea water. Other stressors that are potentially deleterious to S. purpuratus include hypoxia, toxic pollutants, emersion, and rapid burial and smothering in sediment ( i.e. , obrution). Strongylocentrotus purpuratus is resistant to hypoxic conditions (Johansen and Vadas, 1967; Burnett et ah, 2002), but immersion in strongly hypoxic or anoxic lagoon water might have increased the susceptibility of individuals to osmotic stress. Dissolved oxygen (DO) concentration in the lagoon was not recorded during the interval encompassing the mortality events (Abramson, pers. comm.), but it had in the past ranged from < I mg L 1 nearly to saturation (2NDNATURE, 2010). Even so, hypoxic lagoon water would have been aerated while flowing toward the intertidal zone. The lagoon water contains a variety of inorganic and organic pollutants, fecal bacteria, and pathogens (Ambrose and Orme, 2000; Dagit et ah, 2009; BenVau, 201 1), and it is possible that urban runoff from the season’s initial storm contributed a concentrated “first flush” of contaminants to the lagoon. However, toxic contaminants are not seriously elevated at the site (Schiff and Bay, 2003), and a die-off of the fauna within the lagoon did not coincide with the mass mortality of intertidal sea urchins. Therefore it is unlikely that a transitory exposure to pollutants killed S. purpuratus. Conversely, Schroeter (1978) found that multi-day series of minus tides can weaken or kill S. purpuratus through exposure to desiccation, sunlight, and elevated temperatures. Although it is difficult to precisely relate Schroeter’s experimental results to die-offs at Malibu, it is notable that the episode in 2010 was accompanied by a series of minus tides. In contrast, conditions in 2011 may have been relatively less stressful, accounting for the lower mortality of S. purpuratus that year, because the event occurred during plus tides and was preceded by 6 days without minus tides. Following the breaches, some sea urchins may have been killed by rapid burial in sediment, but direct evidence is lacking. Although a considerable amount of sediment from the breach was deposited on the intertidal shelf, waves and currents redistributed it before the reefs were examined. Mass Mortality of S. purpuratus in California The catastrophic natural events, which are reviewed in the Introduction, are not the only agents of massive sea urchin kills. Humans have profoundly altered the population size of S. purpuratus throughout its range, often through their efforts to harvest or to eradicate sea urchins. In that respect, the mass mortalities of S. purpuratus at Malibu Lagoon State Park are exceptional, as they were human-caused but unintentional. They appear to have been “collateral damage” stemming from attempts to improve conditions for recreational surfing. Paleoindians, who were the first Californians to fish for strongylocentrotid sea urchins, unwittingly initiated a 10,000-year interaction with S. purpuratus that has often been detrimental to the species. After Paleoindians overharvested sea otters (Enhvdra lutris (Linnaeus, 1758)) and sheephead wrasse ( Semicossyphus pulcher (Ayres, 1854)), eliminating the predators that had held Strongylocentrotus populations in check, sea urchins became a notable component of the fishers’ diet (Sails, 1995; Erlandson et ah, 2005; Erlandson et ah, 2011). During historical times Californians conspicuously boosted the population growth of S', purpuratus at least twice. In the first instance, well before 1900, sea otters between Alaska and Mexico were hunted almost to extinction by fur traders (Riedman and Estes, 1988), and absent their chief predator, sea urchins are thought to have proliferated 32 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES (Dayton et al.. 1998; Tegner and Dayton, 2000; Foster and Schiel, 1988, 2010). A second population surge of purple sea urchins occurred during the 1950s, when enormous volumes of untreated sewage were discharged by metropolitan Los Angeles and San Diego. Despite the influx of toxic pollutants, sea urchins thrived on the nutrients released at sewage outfalls in southern California (North, 1974). In response to marked increases in the numbers of S. franciscanus and S. purpuratus, concerted efforts have been made to eradicate strongylocentrotids that were deemed undesirable by the kelp-harvesting industry and various fishing interests. Beginning in the 1970s groups of divers armed with hammers smashed sea urchins, and boats spread hundreds of tons of quicklime (calcium oxide) on concentrations of sea urchins, killing them and other fauna. Temporarily exterminating sea urchins may have enhanced the recruitment rate of giant kelp (Macrocystis pyrifera) locally, but widespread recovery of kelp beds did not occur before large-scale upgrades of sewage treatment facilities extensively improved water quality (North, 1974; Foster and Schiel, 2010; Meux and Ford, 2010). A subsistence fishery persisting in the 20th Century among Native American and immigrant populations on the West Coast reduced the numbers of S. purpuratus to some extent (Johnson and Snook, 1935; Greengo, 1952; Heizer and Mills, 1952; Kato and Schroeter, 1985). Some purple sea urchins have also been damaged and removed through human recreational activities (Addessi, 1994; Murray et al., 1999). In contrast, a commercial fishery initiated in the early 1970s had an extensive and long-lasting impact (Kato, 1972; Kalvass and Hendrix, 1997; Parker and Ebert, 2004). In California, old- growth stocks of 5. franciscanus were depleted and populations collapsed, because “until the late 1980’s, the sea urchin fishery was not actively managed...” (Kalvass and Hendrix, 1997:13; Rogers-Bennett, 2007). Fortunately for S. purpuratus , nearly 99% of the commercial harvest consisted of S. franciscanus , and localized removal of red sea urchins may have enhanced the access of purple urchins to food and habitat (Schroeter, 1978; Parker and Ebert, 2004). Conclusions Based on available information, recent mass mortalities of sea urchins at Malibu Lagoon State Beach appear to have been the indirect result of human-caused breaches in the sand bar, a previously unreported source of sea urchin mortality. The presumed cause of death, exposure to low-salinity water, has affected populations of S. purpuratus elsewhere (MacGinitie, 1939; Sousa, 1979; Littler and Littler, 1987; Ebert and Grupe, 2008). At Malibu, S. purpuratus and S. franciscanus seem to have been the only animals killed by the breaches, but that could not be confirmed. Perhaps particular hydrodynamic characteristics of dead purple sea urchins (Denny and Gaylord, 1996; Stewart and Britten-Simmons, 2011) favored their massive accumulation on the tide line. At other localities, freshly killed mollusks and polychaetes have been stranded during storms (Rees et al., 1977), but none washed ashore at Malibu. However, they or other animals that were killed could have been transported offshore, as sometimes occurs following intertidal mass mortalities (e.g., Hendler, 1977; Girard et al., 2012, and pers. comm.). No evidence was found that 5. purpuratus died because of elevated temperature or disease. Hypoxia, hydrogen sulfide, toxic pollutants, or rapid burial in sediment may have played a role in the mortality events, but crucial facts regarding water chemistry and sediment deposition are lacking, as is information regarding the tolerance of adult purple PURPLE SEA URCHIN MASS MORTALITY 33 sea urchins to physical and chemical stressors acting singly and in combination. Thus, a critical understanding of the periodic die-offs at Malibu requires further research on the behavior, physiology, and population dynamics of S. purpuratus, in addition to long-term monitoring of breach incidents and water quality at the lagoon. Mortality events in 2010 and in 2011 occurred in early October, after storms that produced similarly heavy rainfall. Since it is likely that nearly equal volumes of brackish water were released from the lagoon in both events, why were ten times more S. purpuratus killed in 2010? In that year, at low tide, lagoon water flowed directly toward the central depression in Basin Reef, where the seaward edge of the shelf could have deterred marine water from mixing with hyposaline lagoon water and thereby mitigating its harmful effects. Consequently, in 2010, discharge from the lagoon coinciding with a diurnal minus tide, simultaneously exposed S. purpuratus to hyposaline water, sunlight, and dehydration for an extended period. Additionally, after emersion, purple sea urchins replace fluid in their digestive tract with air, using the inflated gut for gas exchange (Burnett et ah, 2002). Having an air-filled gut may have increased their susceptibility to osmotic shock and also made them somewhat buoyant, a possibility which should be experimentally tested. Repeated exposure during several more minus tides may have resulted in the influx of additional dead sea urchins that occurred two days after the breach in 2010. In contrast, in 2011, S. purpuratus was subjected to less stress because Bird Reef did not impound freshwater outflow from the lagoon, and because the series of low tides subsequent to the breach were above datum. Regarding the 201 1 event, timing of the breach relative to the low tide is not known. The paucity of sea urchins found in the mid-intertidal at Malibu Lagoon State Beach was to some extent a result of seagull predation. When the beach was surveyed, western gulls repeatedly were seen attacking sea urchins on the crests of subaerially exposed reefs and around partially submerged cobbles. Their behavior resembled the birds observed at Palos Verdes Peninsula described by Snellen et al. (2007). The authors estimated that one western gull potentially could consume 3,229 S’, purpuratus each year, and that the flock of ~45 birds that they studied could consume 145,305 sea urchins. Since greater numbers of western gulls occur at Malibu, their impact on the sea urchin population could be considerable. Moreover, similarly to the “trophic cascade” created by glaucous-winged gulls (Larus glaucescens Naumann, 1840) in Washington (Wootton, 1997), western gulls might indirectly increase the growth of intertidal algae at Malibu, by reducing the amount of sea urchin herbivory. In the past, severe storms may have periodically devastated the intertidal population of S. purpuratus at Malibu Lagoon, as in 1998 when historic photographs (Schwarz and Orme, 2005: Figs. 6, 7) show that a remarkable El Nino event blanketed the intertidal zone with sand. However, repetitive mass mortalities at Malibu Lagoon, due to natural and artificial breaches and gull predation, appear to maintain the intertidal population of purple sea urchins at consistently low density. Population densities of S. purpuratus vary markedly depending on configuration and composition of the benthic substrate (Sagarin and Gaines, 2002; Ebert, 2010). Nevertheless, density of S', purpuratus at Malibu Lagoon is low compared to other southern California intertidal localities where 20 to 100+ individuals per m2 quite often are found (Schroeter, 1978; Littler and Littler, 1987, Lawrenz-Miller, pers. comm.). Since S. purpuratus is an “ecosystem engineer,” and since its grazing controls algal and invertebrate community structure in southern California (Ebert, 1977; Sousa et al., 1981; Rogers-Bennett, 2007:415), the low population density at Malibu may appreciably affect intertidal ecology of that locale. 34 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES California’s coastal lagoons have been encroached upon and degraded by urban development for over a century, to the point that their restoration to pristine, natural condition generally seems impracticable. Consequently they have been neglected, or engineered to better serve as resources for tourism and recreation and to protect private property and public health. Virtually all efforts to preserve the ecological functionality of lagoons have been directed toward estuarine, freshwater, and terrestrial habitats (2NDNATURE, 2010; Hany and Elwany, 2011; Gladstone et al., 2006; Kraus et ah, 2008). Yet the inland habitats are individual components of an ecosystem that also encompasses the marine environment. The mass mortalities at Malibu Lagoon, described herein, are manifest evidence of a linkage between the lagoon and the nearshore benthic fauna. They are also an indication that the impacts of lagoon breaches on the marine fauna of California, overlooked until now, warrant further study. Acknowledgements This contribution benefited greatly, and indeed would not have been possible, without indispensible advice, information, data, and photographs provided by many generous individuals. I am indebted and grateful to all for their help; Mark Abramson, Santa Monica Bay Restoration Foundation; Tony Barboza, Los Angeles Times; Heather Burdick, Los Angeles Waterkeeper; Rosi Dagit, Resource Conservation District of the Santa Monica Mountains; John Dixon, California Coastal Commission; Thomas Ebert, San Diego State University; John (Jack) Engle, University of California, Santa Barbara; Kimball Garrett, Natural History Museum of Los Angeles County; Timothy Hayden, State Park Ranger; Jim DePompei, Cabrillo Marine Aquarium; Laura Jurgens, University of California, Davis; Ryan Kittell, National Weather Service; Jenna Krug, Resource Conservation District of the Santa Monica Mountains; Suzanne Lawrenz- Miller, Cabrillo Marine Aquarium; Brian Meux, Los Angeles Waterkeeper; Dana Murray, Heal the Bay; James Nebelsick, Tubingen University; David Ogle, Coastal Adventure Charters; Julianne Passarelli, Cabrillo Marine Aquarium; John Pearse, University of California, Santa Cruz; Jan Pechenik, Tufts University; Gregory Pfeifer, Los Angeles County Lifeguard Specialist; Mark Ralston/AFP; Stephen Schroeter, University of California, Santa Barbara; Sarah Sikich, Heal the Bay; Chuck Striplen, San Francisco Estuary Institute; Marshall Thompson, prvideo.tv; Roy van de Hoek, Ballona Institute. In addition, Rosi Dagit, Florence Nishida, John Pearse, Jan Pechenik, Stephen Schroeter, Thomas Ebert, and anonymous reviewers provided valuable comments on drafts of this publication. Literature Cited 2NDNATURE. 2010. Malibu Lagoon restoration monitoring plan (MLRMP) baseline conditions report. Report Submitted to Resource Conservation District of the Santa Monica Mountains. 70 pp. Addessi, L. 1994. Human disturbance and long-term changes on a rocky intertidal community. Ecol. Appl., 4:786-797. Ambrose, R.F. and A.R. Orme. 2000. Lower Malibu Creek and Malibu Lagoon resource enhancement and management. Final Report to the State Coastal Conservancy. University of California, Los Angeles, xvi + 539 pp. Angeloni, A., .1. Bradbury, and A. Chaine. 1999. Growth, seasonality, and dispersion of a population of Aplysia vaccaria Winkler, 1955. Veliger, 42:1-9. Ausderk, T.E. 1979. A field study of growth and reproduction in Aplysia californica. Biol. Bull., 157: 407M-2 1 . BenVau, ,I.M. 2011. Multivariate study of fecal indicator bacteria concentration in the Malibu Lagoon and nearby beaches in Malibu, CA. M.Sc. Thesis, California State University, Chico. 86 pp. PURPLE SEA URCHIN MASS MORTALITY 35 Burnett, L., N. Terwilliger, A. Carroll, D. Jorgensen, and D. Scholnick. 2002. Respiratory and acid-base physiology of the purple sea urchin, Strongylocentrotus purpuratus, during air exposure: presence and function of a facultative lung. Biol. Bull., 203:42-50. Dagit, R.. S. Adams, and S. Drill. 2009. Die off and current status of southern steelhead trout (Oncorhynchus mykiss) in Malibu Creek, Los Angeles County, USA. Bull. So. Calif. Acad. Sci . , 108:1-15. Dayton, P.K. 1975. Experimental evaluation of ecological dominance in a rocky intertidal algal community. Ecol. Monogr., 45:137-159. — . M.J. Tegner, P.B. Edwards, and K..L. Riser. 1998. Sliding baselines, ghosts, and reduced expectations in kelp forest communities. Ecol. Appl., 8:309-322. Denny, M.W., T.L. Daniel, and M.A.R. Koehl. 1985. Mechanical limits to size in wave-swept organisms. Ecol. Monogr., 55:69-102. — and B. Gaylord. 1996. Why the urchin lost its spines: hydrodynamic forces and survivorship in three echinoids. J. Exp. Biol., 199:717-729. Ebeling, A.W., D.R. Laur, and R.J. Rowley. 1985. Severe storm disturbances and reversal of community structure in a southern California kelp forest. Mar. Biol., 84:287-294. Ebert, T.A. 1977. An experimental analysis of sea urchin dynamics and community interactions on a rock jetty. J. Exp. Mar. Biol. Ecol.. 27:1-22. — . 1983. Recruitment in echinoderms. Pp. 169-203 in Echinoderm Studies, Volume 1. M. Jangoux, and J.M. Lawrence (eds.). A. A. Balkema, Rotterdam, vii + 203 pp. — . 2010. Demographic patterns of the purple sea urchin Strongylocentrotus purpuratus along a latitudinal gradient, 1985-1987. Mar. Ecol. Prog. Ser., 406:105-120. — and B.M. Grupe. 2008. Sudden loss of sea urchins at Sunset Bay, Oregon, following documented presence for 40 years. Abstracts, 93,d Ecological Society of America Annual Meeting, Milwaukee. Wisconsin. Available from: http://esameetings.allenpress.com/2008/P10301.HTM via the Internet. Accessed 6 September, 2012. Erlandson, J.M., T.C. Rick, T.J. Braje. M. Casperson, B. Culleton, B. Fulfrost, T. Garcia, D.A. Guthrie, N. Jew, D.J. Kennett, M.L. Moss, L. Reeder, C. Skinner, J. Watts, and L. Willis. 201 1. Paleoindian seafaring, maritime technologies, and coastal foraging on California’s Channel Islands. Science, 331:1181-1185. --. — , J.A. Estes, M.H. Graham, T.J. Braje, and R.L. Vellanoweth. 2005. Sea otters, shellfish, and humans: 10,000 years of ecological interaction on San Miguel Island, California. Pp. 9-21 in Proceedings of the Sixth California Islands Symposium. (D. Garcelon and C. Schwennn, eds.). Institute for Wildlife Studies, Areata, California, xii + 508 pp. Farmanfarmaian, A. and A.C. Giese. 1963. Thermal tolerance and acclimation in the western purple sea urchin. Strongylocentrotus purpuratus. Physiol. Zook. 36:237-243. Foster, M.S. and D.R. Schiel. 1988. Kelp communities and sea otters: keystone species or just another brick in the wall? Pp. 92-1 15 in The Community Ecology of Sea Otters. (G.R. VanBlaricom and J.A. Estes, eds.), Springer-Verlag, Berlin, Germany, xv + 247 pp. — and . 2010. Loss of predators and the collapse of southern California kelp forests (?): alternatives, explanations and generalizations. J. Exp. Mar. Biol. Ecol., 393:59-70. Giese, A.C. and A. Farmanfarmaian. 1963. Resistance of the purple sea urchin to osmotic stress. Biol. Bull., 124:182-192. Gilles, K.W. and J.S. Pearse. 1986. Disease in sea urchins Strongylocentrotus purpuratus : experimental infection and bacterial virulence. Dis. Aquat. Org., 1:105-114. Girard, D.. S. Clemente, K. Toledo-Guedes, A. Brito, and J.C. Hernandez. 2012. A mass mortality of subtropical intertidal populations of the sea urchin Paracentrotus lividus: analysis of potential links with environmental conditions. Mar. Ecol., 33:377-385. Gladstone, W., N. Hacking, and V. Owen. 2006. Effects of artificial openings of intermittently opening estuaries on macroinvertebrate assemblages of the entrance barrier. Estuar. Coast. Shelf Sci., 67: 708-720. Greengo, R.E. 1952. Shellfish foods of the California Indians. Kroeber Anthropol. Soc. Pap., 7:63-114. Hany, M. and M.H.S. Elwany. 201 1. Characteristics, restoration and enhancement of Southern California lagoons. J. Coastal Res., Special Issue, 59:246-255. Hedgpeth, J.W. and J..I. Gonor. 1969. Aspects of the potential effect of thermal alteration on marine and estuarine benthos Pp. 80-1 15 in Biological Aspects of Thermal Pollution. (P.A. Krenkel and F.L. Parker, eds.), Vanderbilt University Press, Nashville, Tennessee, xx + 407 pp. 36 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Heizer, R.F. and J.E. Mills. 1952. The Four Ages of Tsurai: A Documentary History of the Indian Village on Trinidad Bay. University of California Press, Berkeley and Los Angeles. 207 pp. Hendler, G. 1977. The differential effects of seasonal stress and predation on the stability of reef-flat echinoid populations. Pp. 217-223 in Proceedings of the Third International Coral Reef Symposium, Vol. 1. (D.L. Taylor, ed.). Rosenstiel School of Marine and Atmospheric Science, Miami, Florida, xxiv + 636 pp. Johansen, K. and R.L. Vadas. 1967. Oxygen uptake and responses to respiratory stress in sea urchins. Biol. Bull., 132:16-22. Johnson, M.E. and H.J. Snook. 1935. Seashore Animals of the Pacific Coast. MacMillan, New York. 659 pp. Johnson, P.T. 1971. Studies on diseased urchins from Pi. Loma. Pp. 82-90 in Kelp Habitat Improvement Project, Annual Report 1970-1971. (W.J. North, ed.), California Institute of Technology, Pasadena, California. 150 pp. Kalvass, P.E. and ,1 M. Hendrix. 1997. The California red sea urchin, Strongylocentrotus franciscanus, fishery: catch, effort, and management tools. Mar. Fish. Rev., 59:1 17. Kato, S. 1972. Sea urchins: A new fishery develops in California. Mar. Fish. Rev., 34:23-30. and S.C. Schroeter. 1985. Biology of the red sea urchin, Strongylocentrotus franciscanus , and its fishery in California. Mar. Fish. Rev., 47:1-20. Kraus, N.C., K. Putsch, and S. M linger. 2008. Barrier beach breaching from the lagoon side, with reference to Northern California. Shore & Beach, 76:33-43. Lambert, P. and W.C. Austin. 2007. Brittle Stars, Sea Urchins and Feather Stars of British Columbia, Southeast Alaska and Puget Sound. Royal BC Museum, Victoria. 150 pp. Lester, S.E., E.D. Tobin, and M.D. Behrens. 2007. Disease dynamics and the potential role of thermal stress in the sea urchin, Strongylocentrotus purpuratus. Can. J. Fish. Aquat. Sci., 64:314-323. Littler, M.M. and D.S. Littler. 1987. Effects of stochastic processes on rocky-intertidal biotas: an unusual flash flood near Corona del Mar, California. Bull. So. Calif. Acad. Sci., 86:95-106. MacGinitie, G.E. 1939. Some effects of fresh water on the fauna of a marine harbor. Am. Midi. Nat., 21: 681-686. Meux, B. and T. Ford. 2010. Giant kelp community restoration in Santa Monica Bay. Urban Coast, 2:43-46. Murray, S., T.G. Denis, J.S. Kido, and J.R. Smith. 1999. Human visitation and the frequency and potential effects of collecting on rocky intertidal populations in southern California marine reserves. CalCOFI Rep., 40:100-106. North, W.J. 1974. A review of studies supporting sea urchin control as a means of restoring kelp beds. Pp. 95-108 in Kelp Habitat Improvement Project, Annual Report 1973-1974. (W.J. North, ed.), California Institute of Technology, Pasadena, California. 137 pp. Paine. R.T. and R.L. Vadas. 1969. The effects of grazing by sea urchins, Strongylocentrotus spp., on benthic algal populations. Limnol. Oceanog., 14:710-719. Parker, D.O. and T.A. Ebert. 2004. Purple sea urchin. Pp. 10-1-10-4 in Annual Status of the Fisheries. Report through 2003. (C. Ryan and M. Patytren, eds.), California Department of Fish and Game, vi + 161 pp. Pearse, J.S.. D P. Costa, M.B. Yellin. and C.R. Agegian. 1977. Localized mass mortality of red sea urchin, Strongylocentrotus franciscanus near Santa Cruz, California. U.S. Fish. Bull., 75:645-648. — and A.H. Hines. 1987. Long-term population dynamics of sea urchins in a central California kelp forest: rare recruitment and rapid decline. Mar. Ecol. Prog. Ser., 39:275-283. — and R. Mooi. 2007. Echinoidea. Pp. 914-922 in The Light and Smith Manual: Intertidal Invertebrates of Central California and Oregon, Fourth Edition. (J.T. Carlton, ed.). University of California Press, Berkeley, xvii + 1001 pp. — and V.B. Pearse. 1975. Growth zones in the echinoid skeleton. Amer. Zook. 15:731-753. Rees, E.I.S., A. Nicholaidou, and P. Laskaridou. 1977. The effects of storms on the dynamics of shallow water benthic associations. Pp. 465^-73 in Biology of Benthic Organisms. 1 1 th European Symposium on Marine Biology, Galway, October 1976. (B.F. Keegan, P.O. Ceidigh, and P.J.S. Boaden, eds.), Pergamon Press. Oxford, xxxiii + 630 pp. Riedel, B., M. Zuschin. and M. Stachowitsch. 2012. Tolerance of benthic macrofauna to hypoxia and anoxia in shallow coastal seas: a realistic scenario. Mar. Ecol. Prog. Ser., 458:39-52. Riedman, M.L. and J.A. Estes. 1988. A review of the history, distribution and foraging ecology of sea otters. Pp. 4-21 in The Community Ecology of Sea Otters. (G.R. VanBlaricom and J.A. Estes, eds.). Springer- Verlag, Berlin, Germany, xv + 247 pp. PURPLE SEA URCHIN MASS MORTALITY 37 Rogers-Bennett, L. 2007. The ecology of Strongylocentrotus franciscanus and Strongylocentrotus purpuratus. Pp. 393^125 in Edible Sea Urchins: Biology and Ecology. (J.M. Lawrence, ed.), Elsevier, Amsterdam, The Netherlands, xxiii + 529 pp. Rowley, R.J. 1990. Newly settled sea urchins in a kelp bed and urchin barren ground: a comparison of growth and mortality. Mar. Ecol. Prog. Ser., 62:229-240. Sagarin, R D and S.D. Gaines. 2002. Geographical abundance distributions of coastal invertebrates: using one-dimensional ranges to test biogeographic hypotheses. J. Biogeogr., 29:985-997. Sails, R.A. 1995. Ten-thousand years of fishing; the evidence for alternate stable states in nearshore ecosystems as the result of overexploitation of the California sheephead ( Semiseossyphus pulcher) by prehistoric fishermen on San Clemente Island, California. Pp. 204-214 in Memorias IX Simposium Internacional Biologia Marina (June 1992) Universidad Autonoma Baja California Sur. La Paz, Baja California Sur, Mexico, 275 pp. Schiff, K. and S. Bay. 2003. Impacts of stormwater discharges on the nearshore benthic environment of Santa Monica Bay. Mar. Environ. Res., 56:225-243. Schroeter, S.C. 1978. Experimental studies of competition as a factor affecting the distribution and abundance of purple sea urchins, Strongylocentrotus purpuratus (Stimpson). Ph.D. Dissertation, University of California, Santa Barbara, xi + 184 pp. Schwarz, K.M. and A.R. Orme. 2005. Opening and closure of a seasonal river mouth: The Malibu estuary- barrier-lagoon system, California. Z. Geomorphol., Suppl., 141:91 -109. Snellen, C.L.. P.J. Hodum, and E. Fernandez-Juricic. 2007. Assessing western gull predation on purple sea urchins in the rocky intertidal using optimal foraging theory. Can. J. Zool., 85:221-231. Sousa, W.P. 1979. Experimental investigations of disturbance and ecological succession in a rocky intertidal algal community. Ecol. Monogr., 49:227-254. — , S.C. Schroeter, and S.D. Gaines. 1981. Latitudinal variation in intertidal algal community structure: the influence of grazing and vegetation propagation. Oecologia (Berk), 48:297-307. Stewart, I LL. and K.H. Britton-Simmons. 201 1. Streamlining behaviour of the red urchin Strongylocen- trotus franciscanus in response to llow. J. Exp. Biol., 214:2655-2659. Swift, C.C., T.R. Haglund, M. Ruiz, and R.N. Fisher. 1993. The status and distribution of the freshwater fishes of southern California. Bulk So. Calif. Acad. Sci., 92:101 167. Tegner, M.J. and P.K.. Dayton. 1981. Population structure, recruitment and mortality of two sea urchins (Strongylocentrotus franciscanus and S. purpuratus) in a kelp forest. Mar. Ecol. Prog. Ser., 5: 255-268. — and . 1991. Sea urchins, El Ninos, and the long term stability of Southern California kelp forest communities. Mar. Ecol. Prog. Ser., 77:49-63. — and . 2000. Ecosystem effects of fishing in kelp forest communities. ICES J. Mar. Sci., 57: 579-589. Thompson. B., S. Bay, D. Greenstein, and J. Laughlin. 1991. Sublethal effects of hydrogen sulfide in sediments on the urchin Lytechinus pictus. Mar. Environ. Res., 31:309-321. Ward, J.R. and K.D. Lafferty. 2004. The elusive baseline of marine disease: Are diseases in ocean ecosystems increasing? PLoS Biol., 2:0542-0547. Wootton. T.J. 1997. Estimates and tests of per capita interaction strength: diet, abundance, and impact of intertidally foraging birds. Ecol. Monogr., 67:45-64. Bull. Southern California Acad. Sci. 112(1), 2013, pp. 38-48 © Southern California Academy of Sciences, 2013 Review of the Purple Amole Chlorogalum purpureum (Agavaceae): a Threatened Plant in the Coast Ranges of Central California Christopher P. Kofron,1* Connie Rutherford,1 Elizabeth R. Clark,2 Darlene Woodbury,3 Jody Olson,4 and Robert F. Holland5 1 U. S. Fish and Wildlife Service. 2493 Portola Road. Suite B. Ventura, CA 93003 2 Environmental Division, U.S. Army Garrison Fort Hunter Liggett, Fort Hunter Liggett, CA 93928 'Center for Environmental Management of Military Lands, Colorado State University, Fort Hunter Liggett, CA 93928 4 Center for Environmental Management of Military Lands, Colorado State University, Building 910, Camp Roberts, CA 93451 5 Geobotanical Phenomenology, 3371 Ayres Holmes Road, Auburn, CA 95602 Abstract. — The purple amole Chlorogalum purpureum (Agavaceae) is a bulbous, perennial soap plant endemic to central California and listed as threatened under the U.S. Endangered Species Act since 2000. Chlorogalum p. purpureum occurs in the rain shadow of the Santa Lucia Range on Fort Hunter Liggett, south Monterey Co., and on Camp Roberts, north San Luis Obispo Co. Chlorogalum p. reduction occurs in the rain shadow of the La Panza Range in central San Luis Obispo Co., mostly on Los Padres National Forest and with potential for a substantially larger occupied area on private land. We review and enhance the existing knowledge of C. purpureum , in particular its life history and ecology, distribution, population sizes, threats, current management and conservation status. In 2012, invasive plants are the primary threat to C. purpureum. Introduction The genus Chlorogalum (Agavaceae, Bolger et al. 2006) comprises five plant species inhabiting western North America from southern Oregon, USA, to Baja California, Mexico (Jernstedt 2012). One species, purple amole Chlorogalum purpureum , is listed as threatened under the U.S. Endangered Species Act (U.S. Fish and Wildlife Service [USFWS] 2000). Chlorogalum purpureum is a bulbous, perennial soap plant and the only member of the genus with purple or blue flowers, the other species having white or pink flowers (Hoover 1940, Jernstedt 2012). It is endemic to central California (Figure 1 ), with two varieties recognized (Hoover 1964): purple amole C. purpureum purpureum (25 to 40 cm tall), and Camatta Canyon amole C. purpureum reduction (10 to 20 cm tall, Jernstedt 2012). At Federal listing in 2000, C. p. purpureum was known from three occurrences on Fort Hunter Liggett (a training installation of the U.S. Army; 35°54'4.37"N, 121°10'40.33"W), Monterey Co., with an estimated population of ~ 13,450 plants (USFWS 2000). Primary threats were loss of plants and habitat by military activities, and displacement by invasive grasses. In 2000 after listing, a population estimated at 10,000 plants was discovered on * Corresponding author: chris_kofron@fws.gov 38 REVIEW OF THE THREATENED PURPLE AMOLE CHLOROGALUM PURPUREUM 39 Camp Roberts (a training installation of the California Army National Guard; 35°45'35.29"N, 120°50'3.91"W), San Luis Obispo Co. (USFWS 2001), which was estimated in 2001 to comprise > 200,000 plants (California Army National Guard in USFWS 2002). Chlorogalum p. reductum was known from three occurrences in the La Panza region in central San Luis Obispo Co. at listing: on Los Padres National Forest (35°24'2.69"N, 120°16'49.01"W), and two nearby private properties. Population estimates ranged from ~ 1,300 to 300,000 plants on ~ 4 ha. Primary threats were illegal vehicle trespass on national forest land, road maintenance, displacement by invasive grasses, and cattle grazing (USFWS 2000). The State of California had previously listed C. purpureum reductum as rare under the California Endangered Species Act in 1978 (California Department of Fish and Game 2012b). Our purpose is to review and enhance the existing knowledge of C. purpureum , in particular its life history and ecology, distribution, population sizes, threats, current management and conservation status. Methods In the literature, the common name purple amole lias been used for the species C. purpureum and also for the nominate variety. To avoid confusion, hereafter we use purple amole only for C. p. purpureum. Throughout this paper, “we” refers to all or any one of the authors. We summarize the knowledge of the species, including current threats and management, and analyze the conservation status in 2012. The expressed views are solely ours and not the official position of any agency. We managed purple amole on Fort Hunter Liggett (66,773 ha) from 1998 to 2012 and on Camp Roberts (17,314 ha) from 2001 to 2012. At Fort Hunter Liggett, we have maintained a database since 2000 with attributes of the population that we update as new information is obtained by annual monitoring of life history parameters and by additional analyses. We mapped 880 patches from 1997 to 2012, established transects in some patches, mapped and numbered (marked) some individual plants, and estimated the number of plants in most patches: 1 to 10, 11 to 50, 51 to 100, 101 to 200, 201 to 500, 501 to 1,000, 1,001 to 2,000, 2,001 to 5,000 or > 5,000. At Camp Roberts, we censused the population during May to June 2001 by demarcating the occupied area into quadrats (each 25 x 100 m) and then counting (using a clicker counter) the number of plants in each quadrat, stopping at 5,001 (> 5,000). Our counts are only for plants with above- ground structures, which excludes the bulbous stage. For Camatta Canyon amole, we conducted three site visits (2010, 2011, 2012), communicated with relevant persons and examined soil maps. In addition, we reviewed the literature. We consider a location as a separate occurrence only if it is > 0.4 km from the nearest occurrence (California Department of Fish and Game 2012a). Review of the Species The species inhabits a semiarid environment with hot dry summers and cool wet winters. Both varieties grow in gravelly clay soil in open areas with a light cover of native plants in grassland, blue oak savanna (Quercus douglasii ) and blue oak woodland. Records for purple amole account for 97% (~ 429 ha) of the known occupied area, and for Camatta Canyon amole 3% (~ 12 ha, our current estimate). Both varieties sometimes grow in association with cryptogamic crusts (cyanobacteria, lichens, mosses and fungi on the soil surface; E.L. Painter in USFWS 2000, Guretzky et al. 2005, pers. obs.). Cryptogamic crusts are important elements of arid and semiarid ecosystems (Beymer and 40 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES dots, purple amole C. p. purpureum; black squares, Camatta Canyon arnole C. purpureum reduction. Klopatek 1992) because they stabilize soil against erosion, fix atmospheric nitrogen, form organic matter (Eldridge and Greene 1994), retain soil moisture, discourage weed growth (Belnap et al. 2001) and provide favorable sites for growth of native plants (e.g., Lesica and Shelly 1992). New C. purpureum become established almost entirely by seed (Hoover 1940). Purple amole in a screen house llowered 3 to 4 y after planting as seed, and most died 5 to 7 y after germination (D. Wilken, Santa Barbara Botanic Garden, California, pers. comm.). Camatta Canyon amole planted on site as seed required at least 12 y to Dower (Koch and Hillyard 2009). Purple amole emerge above ground Nov. to Dec., with inDorescences developing during early spring, and Dower and fruit during May to June. Camatta Canyon amole Dower and fruit during Apr. to June. As fruits mature, the leaves wither and inDorescences dry and turn brown. We observed purple amole bulbs entering dormancy during summer, and some marked plants remained dormant for multiple years. Purple Amole Purple amole is now known to occur on two properties: 17 occurrences on Fort Hunter Liggett, and one occurrence on Camp Roberts (Figure 1). It is endemic to the Santa Lucia Range in south Monterey Co. and north San Luis Obispo Co., occurring in the rain shadow at lower elevations (213 to 390 m) on the east side. The Coast Ridge rises to 1,000 m west of Fort Hunter Liggett, and Pine Mountain rises to 1,095 m west of Camp Roberts. Mean annual rainfall is 485 mm at Fort Hunter Liggett (Cantonment, primarily Nov. to Apr.; National Park Service 2007) and 285 mm at Camp Roberts (primarily Oct. to May, White et al. 2000). All known occurrences are in the San Antonio and Nacimiento River valleys. Construction of Lake Nacimiento and Lake San Antonio in 1961 and 1965, respectively, may have submerged some occupied areas. REVIEW OF THE THREATENED PURPLE AMOLE CHLOROGALUM PURPUREUM 41 The land comprising Fort Hunter Liggett has been settled and used by EuroAmericans since 1771, which has affected current distribution. Purple amole is nearly absent from areas previously ripped/disked for cultivation, and most abundant in areas not previously ripped/disked including some used for intensive military training since 1941. The taxon occupies ~ 341 ha on Fort Hunter Liggett, occurring in hundreds of discontinuous patches at mean density up to 1 1 plants/m2 in grassland, blue oak savanna and open areas of blue oak woodland. Using our cumulative data since 1997, we estimate the population comprises 315,000 to 700,000 plants. Records on Fort Hunter Liggett are at 270 to 390 m elevation. Surveys along the east boundary suggest it likely occurs also on adjacent private land (Wilken in USFWS 2002), for which the USFWS (2002, 2003) designated critical habitat (620 ha; an area with essential characteristics). Purple amole on Fort Hunter Liggett occurs on both deep and thin soils, most of which are loamy and underlain by clay with fine gravel (generally < 0.5 cm diameter) on the surface. Most occupied areas (78%) are level or near-level (slope < 10%), including a few on ridge top terraces (Wilken in USFWS 2001 ). It generally grows where other vegetation is sparse (Guretzky et al. 2005) and in association with low-growing annual plants, including: natives — rusty popcorn (lower Plagiobothrys nothofulvus, miniature lupine Lupimis bicolor and California goldfields Lasthenia californica; and invasives — soft brome Browns hordeaceus and stork’s bill Erodium sp. Purple amole and cryptogamic crusts co- occur in 50% of the survey plots, and density of purple amole is positively correlated (P < 0.05) with cryptogamic crusts (r = 0.271) and other native plants (r = 0.199, Guretzky et al. 2005). Niceswanger (2002) reported Botta’s pocket gophers Thomomys bottae eating purple amole (stems, leaves, flowers) on Fort Hunter Liggett. We observed Botta’s pocket gophers displacing bulbs and causing extensive disturbance to the habitat. Since 2010, feral pigs Sus scrofa are also disturbing the habitat. In 2003, the U.S. Army used a ball and chain to clear chaparral (Kucera and Mayer 1999) at several sites presumably not occupied by purple amole. However, one of these sites (~ 78 ha) now likely has the highest density on Fort Hunter Liggett. At another site, fire burned a study plot before seed maturation, with apparent loss of all seeds (Niceswanger 2002). The following year, the number of plants at the burned site increased by 175%. These observations suggest purple amole may respond favorably to removal of potentially competitive species (invasives and natives) and to fire at certain times of the year. We are testing the potential of prescribed fire to remove invasive plants and benefit purple amole, with several areas already burned in 201 1. Over 7 y at Fort Hunter Liggett, we observed 26% of monitored plants (range 13 to 63% per y) attempting reproduction (produced stalks, flowers, capsules or seeds), with 12% (range 5 to 26% per y) actually producing seeds. Flowering was correlated with total rainfall in the preceding few months (Guretzky et al. 2005). Numbers of seeds ranged up to 386 per plant (mean = 28, Niceswanger 2002). We observed many marked plants in dormancy as bulbs for 1 to 3 y, and one marked plant remained dormant for 6 y and another 5 y. Guretzky et al. (2005) estimated the dormancy rate as 23% of the population per y and the mortality rate as 10% of the population per y. Since Federal listing, the U.S. Army has included purple amole in its natural resources management plan at Fort Hunter Liggett and annually monitors life history parameters. In 2005, 2007 and 2010, the U.S. Army consulted with the USFWS regarding its ongoing activities and effects on purple amole. In accordance with biological opinions issued under the U.S. Endangered Species Act, cumulative loss of occupied areas by ground disturbance does not exceed 0.4 ha/y. The U.S. Army terminated cattle grazing on Fort 42 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Hunter Liggett in 1991 because of adverse effects to natural resources, and it is not likely to resume in the occupied areas. Although the population appears stable on Fort Hunter Liggett, we identify invasive plants, Botta’s pocket gophers, feral pigs and possibly lack of fire as threats in 2012 (Table 1). Purple amole on Camp Roberts occupies ~ 88 ha north of the Nacimiento River on the west side of the installation (213 to 274 m elevation), which is ~ 31 km from the nearest occurrence on Fort Hunter Liggett. The most-recent census in 2001 recorded > 250,000 plants (our refined analysis). In June 2005, a hot wildfire burned ~ 90% of the occupied area with loss of the seed crop. A dense layer of thatch (10 to 20 cm) had accumulated since cessation of sheep grazing. In 2006, we observed 10.4 plants/m2 in survey plots during Mar. and 5.6 plants/m2 during May. The decrease likely represented some plants entering dormancy without flowering. Although not tested and anecdotal, plants in the burned area appeared taller and with increased reproductive vigor (more flowers, fruits or seeds) than plants in the non-burned area, which suggests purple amole may respond favorably to fire at certain times of the year. On Camp Roberts, purple amole occurs predominately in soil with a high con- centration of gravel underlain by hard-packed clay, growing across the fiat surface of an old alluvial fan primarily on northeast exposures with slope < 10%. The vegetation is predominantly grassland, with blue oak woodland on surrounding slopes. Here purple amole grows where there is sparse plant cover, seldom under oak canopies, and usually among low-growing native grasses (especially small fescue Vulpia microstachys ) and herbaceous native plants. In particular, yellowfiower tarweed Holocarpha virgata co- occurred in 95% of survey plots, annual agoseris Agoseris heterophylla in 82%, and bigflower agoseris Agoseris grandiflora in 79%. Together, yellowfiower tarweed and bigflower agoseris comprised 75% of total plant cover in survey plots with purple amole. Invasive plants in the habitat include wild oat Arena fatua , soft brome and redstem stork’s bill Erodium cicutarium. Feral pigs eat the bulbs of purple amole on Camp Roberts and previously caused extensive disturbance to the habitat. This threat has been reduced by controlled hunting elsewhere on the installation. The California Army National Guard ceased training activities in areas occupied by purple amole in 2000 (except occasional road use), along with sheep grazing and hunting. The agency included the taxon in its natural resources management plan for Camp Roberts and annually monitors life history parameters. The California Army National Guard consults with the USFWS regarding its activities and effects on purple amole. Consequently, limited military activities resumed in occupied areas in 2011. We identify the following threats on Camp Roberts in 2012: invasive plants, feral pigs and possibly lack of fire (Table 1). We are considering several actions that may benefit purple amole, including prescribed fire and controlled sheep grazing to remove invasive plants. Camatta Canyon Aniole Camatta Canyon amole is now known from four occurrences on four properties (Figure 1 ), ~ 61 km southeast of purple amole on Camp Roberts and ~ 92 km southeast of purple amole on Fort Hunter Liggett. It is endemic to the La Panza Range in central San Luis Obispo Co., occurring in the rain shadow at intermediate elevations (570 to 633 m) on the east side. Black Mountain rises to 1.104 m immediately to the west. Based on data for La Panza Ranch (~ 10 km eastward and therefore drier), mean annual rainfall at the occupied area is > 156 mm (likely by several cm), along with rain primarily from Oct. to Apr. (http://www.worldclimate.com 2011). Table 1. Status of the threatened Chlorogalum purpureum (Agavaceae) at listing in 2000 and since listing to 2012. Invasive plants are now the primary threat. CR denotes Camp Roberts. FHL denotes Fort Hunter Liggett. LPNF denotes Los Padres National Forest. USFWS denotes U.S. Fish and Wildlife Service. REVIEW OF THE THREATENED PURPLE AMOLE CHLOROGALUM PURPUREUM 43 U U o E U a u fN ° Cl ^ rn O ^ O o o o D- o' ON © hJ I pl oc u o_ © ►J 1 Cl _ g Pi O u - ' _1 2 jl o X 2 Cl r- < c C O P z ■ U5 E "O C/3 jf* IS W ° ^ ,v d u u bO .£ 5 o t> b£» & X Ph C/3 £ Ph ^ 3 C/3 - £ 2 = =? 5H.I | „ a. C (L) C ■2 e y bfj o B -0 V. 7, -o y o Oi C£ Oi - T X I Jh Ph p_ p- LL p_ _ CP ' p . bO bij QJ <3 .S .S O ‘*3 O o ^ c M M S"£ C C g o O O vi o- "2 o C i- s 2 C Cl Cl Z Z CL, CL _l I o o cc u _i J 5 I I d Cl PL j M M w .S .E i o o t bX) bO ^ C C CL) o o z bp 'a cb O .= .S G Q. <- S N O c CL) C3 > C "O 'O g E7 > -.5 ^ ° •- g Cg G (L) ^ U I < u bo t: -E C3 +— < CP .til C3 & 7 I * bp IS X5 c: o O 44 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES The primary occupied area is on a ridge top ~ 200 m east of the south end of Canratta Canyon, which is immediately south of the junction of State Highway 58 and Red Hill Road. The population as currently known is mostly on Los Padres National Forest, extending also onto both sides of State Highway 58 in the right-of-way (1.2 m beyond each shoulder), north onto private property (also likely east), and 3.3 km south of State Highway 58 on private property. The right-of-way of State Highway 58 is managed by the California Department of Transportation, which designated the right-of-way as a botanical management area. This agency gives greater environmental review when planning work here (T. Edell, California Department of Transportation, pers. comm.). The private property immediately north of State Highway 58 is a cattle ranch. The private property south of State Highway 58 appears to be a residential property, and it was registered with The Nature Conservancy in a land protection program (California Department of Fish and Game 2007). The extent of the population across the four properties is not precisely known. However, the occupied area was estimated to comprise 3.2 ha on national forest land, 0.1 ha on the private land south of State Highway 58 (USFWS 2000, 2002), and 0.5 ha on the right-of-way of State Highway 58 (California Department of Fish and Game 2007). Records of the California Department of Fish and Game (2007) state the occupied area comprises 51 ha. Abundance estimates for Camatta Canyon amole on national forest land have ranged from tens of thousands to ~ 500,000 plants (California Department of Fish and Game 2007). The private property south of State Highway 58 contained an estimated several hundred plants (A. Koch in USFWS 2002). In the right-of-way of State Highway 58, 213 and 306 plants were observed in 2000 (J. Luchetta in USFWS 2001) and 2005 (T. Edell, California Department of Transportation, pers. comm.), respectively. On the private property north of State Highway 58, we were able to observe Camatta Canyon amole only from the right-of-way in May 2011. We counted ~ 60 plants from one strategic point, observing predominantly tall grasses in the habitat and an absence of grazing. Mean densities in 10 random plots on national forest land in 1987 and 1988 were 6.5 plants/m2 and 10.7 plants/m2, respectively (Magney 1988). During May 2011, we observed a general absence of Camatta Canyon amole among taller invasive plants (~ 30 to 46 cm height) that were especially dense in vicinity of blue oak trees. In contrast, a previous staging area for off-highway vehicles (~ 929 nr) with relatively few invasive plants contained 3.2 Camatta Canyon amole/nr. In addition, we observed Botta’s pocket gophers causing extensive disturbance to the habitat in 2011 and 2012, especially where invasive plants were abundant. In May 2012, the mean density of Camatta Canyon amole in the most densely populated part of the occupied habitat (18 quadrats, each 0.5 X 50 m) was 5.3 plants/m2 (L. Simpson, U.S. Forest Service, pers. comm.). Some authors reported Camatta Canyon amole on serpentine soil (e.g.. Hoover 1964), with Safford et al. (2005) reporting it as a strict endemic. We correct and clarify by stating that Camatta Canyon amole on Los Padres National Forest occurs on an alluvial terrace in soil with a red sandy matrix dominated by gravel, all derived from sandstone and shale with no serpentine (D. Chipping, California Polytechnic State University, pers. comm.; Natural Resources Conservation Service 2003). In the vicinity of Red Hill Road, this particular soil type (Arbuckle sandy loam with slope < 10%) comprises ~ 11 ha on Los Padres National Forest and ~ 96 ha on private land immediately north of State Highway 58 (http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx, 2010). Potential exists for the occupied area on private land immediately north to be substantially larger than on REVIEW OF THE THREATENED PURPLE AMOLE CHLOROGALUXl PURPUREUM 45 national forest land. The USFWS (2002, 2003) designated critical habitat for Camatta Canyon amole on 1,324 ha of private land immediately north. On Los Padres National Forest, Camatta Canyon amole grows in open areas predominantly in annual grassland and blue oak savanna, and often co-occurring with cryptogamic crusts. Native plant associates include crown brodiaea Brodiaea coronaria , winecup clarkia Clarkia purpurea , sand pygmy weed Crassula erect a, bluedicks Dichelostemma capitatum , sanicle Sanicula spp., California goldfields, sky lupine Lupinus nanus and gilia Gilia sp. (Magney 1988, Service 2002). Along State Highway 58, it typically grows in hard-packed soil and away from dense grasses, either among low- growing plants or in open areas around chamise Adenostoma fasciculatum (T. Edell. California Department of Transportation, pers. comm.). The occupied area on national forest land is in the Pozo-La Panza Unit, which is best known for off-highway vehicle recreation, and it is bisected by a dirt road. In 2010, 2011 and 2012, we observed that the occupied area was effectively protected from off-highway vehicles by fencing, signs and ranger patrols. The adverse effects of road maintenance (in particular grading) appeared to be ongoing, but which we consider a lesser threat. It is also in the Navajo Allotment where a permittee usually grazes cattle from Feb. to May, which overlaps in time with flowering and fruiting of Camatta Canyon amole (Apr. to June). Cattle can cause physical damage by trampling, along with soil compaction and erosion, damage to cryptogamic crusts, reducing presence of native plants and increasing presence of invasives (Fleischner 1994, DiTomaso 2000, Belnap and Eldridge 2001). In 2010, we observed the relatively flat soil surface to be imprinted with hoof depressions where cattle had been when the soil was soft and wet. In 2011 and 2012, the allotment had not been grazed. In each year we observed potential competition with invasive plants (e.g., red brome Bromus rubens , slender oat Arena barbata , soft brome) and also possibly native chamise, along with buildup of thatch. Germane et al. (2001) previously recommended grazing as a tool for managing invasive grasses and helping to conserve declining native species in California. In consideration of all factors, controlled cattle grazing at the right density and timing may benefit Camatta Canyon amole by reducing the presence of invasive plants and thatch. In summary, we identify the following threats to Camatta Canyon amole in 2012: invasive plants, Botta’s pocket gophers, uncontrolled cattle grazing, road maintenance, and possibly lack of fire (Table 1 ). The USFWS and U.S. Forest Service are discussing management of Camatta Canyon amole on national forest land. Conservation Status The LISFWS (2008) reviewed the status of the species in 2008 and recommended no change in its listing status of threatened. Although more information is now available, a five-factor analysis using USFWS criteria is beyond the scope of this paper. Nonetheless, using the best available information and international standards (IUCN 2001, 2010), C. purpureum in 201 1 does not meet IUCN criteria for placement on its red list as critically endangered, endangered or vulnerable: extent of occurrence, 931 km"; area of occupancy, 4.41 km2; quality of the habitat, declining; and number of occurrences, 22. Nor does C. p. purpureum meet IUCN criteria for listing: extent of occurrence, 310 km-; area of occupancy, 4.29 km"; quality of the habitat, declining; and number of occurrences, 18. However, C. purpureum reduction meets IUCN criteria for placement on its red list as endangered: extent of occurrence, 2.3 km2; area of occupancy, 0.12 km2; quality of the habitat, declining; and number of occurrences, 4. With these attributes, C. purpureum reduction faces a very high risk of extinction. 46 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Conclusions Since Federal listing in 2000, purple amole is now known to occur in substantially greater numbers, occupy a substantially greater area and occur on two properties, which are managed as separate populations: 17 occurrences on Fort Hunter Liggett, and one on Camp Roberts (Table 1). Unknown occupied areas may possibly occur in the San Antonio and Nacimiento river valleys between the two populations, especially in designated critical habitat (USFWS 2002, 2003). Camatta Canyon amole is known from only four occurrences within a small area: mostly on Los Padres National Forest, two private properties, and the right-of-way of State Highway 58 at Red Hill Road. Based on landscape characteristics (Arbuckle sandy loam and slope < 10%), potential exists for a substantially larger occupied area on private land north of State Highway 58, especially in designated critical habitat (USFWS 2002, 2003). The U.S. Endangered Species Act is the primary Federal law protecting the species, although it has limited ability to protect listed plants on private land. Camatta Canyon amole is also listed under the California Endangered Species Act (California Department of Fish and Game 2007); however, grazing and other agricultural activities on private land are exempt. Invasive plants are now the primary threat to C. purpureum. Invasives may be able to displace it by outcompeting and monopolizing limited resources (e.g., space, sunlight, nutrients, water; Stephenson and Calcarone 1999), and by producing thatch, which can prevent growth and recruitment and also alter the natural fire regime (Brooks et al. 2004). In addition, based on our observations on Fort Hunter Liggett and Los Padres National Forest, some native plants (climax chaparral) may be able to displace C. purpureum , possibly in the absence of fire. We are considering several management actions that may benefit purple amole on Fort Hunter Liggett (prescribed fire) and Camp Roberts (prescribed fire, controlled sheep grazing) by removing invasive plants and thatch. Also, the species may respond favorably to fire in ways not yet known. For example, the wavyleaf soap plant C. pomeridianum exhibits fire-stimulated flowering and produces seeds only in the first year after fire (Borchert and Tyler 2009). However, caution must be exercised because fire can destroy the flowers and seeds of purple amole and also benefit some invasive plants (Klinger et al. 2006). We encourage controlled cattle grazing on Los Padres National Forest and the relevant private land to remove invasive plants and reduce thatch, so long as consideration is given to stock density, the life cycle of Camatta Canyon amole and soil conditions. The effects of the grazing regime should be thoroughly considered. Acknowledgements We acknowledge Dieter Wilken and JoAnn Froland for leading the census at Camp Roberts in 2001. We thank the following for various contributions: Angela Chapman, David Chipping, Douglass Cooper, Kevin Cooper, Nancy Hill Cunha, Thomas Edell, Gary Houston, Tamara Kleeman, Fletcher Linton, Justin Maciulis, Jenness McBride, Gary Montgomery, David Pereksta, Roger Root, Lloyd Simpson and Kirk Wain. Without the assistance of all these persons, this project could not have been accomplished. Literature Cited Belnap, J. and D. Eldridge. 2001. Disturbance and recovery of biological soil crusts. Pp. 363-383 In: Biological Soil Crusts: Structure, Function, and Management. (Belnap, J. and O.L. Lange, eds.) Springer-Verlag, Berlin. 503 pp. REVIEW OF THE THREATENED PURPLE AMOLE CHLOROGALUM PURPUREUM 47 — . J.H. Kaltenecker. R. Rosentreter, J. Williams, S, Leonard, and D. Eldridge. 2001. Biological Soil Crusts: Ecology and Management. Technical Reference 1730-2. Bureau of Land Management, Denver. 1 10 pp. Beymer, R.J. and J.M. Klopatek. 1992. Effects of grazing on cryptogamic crusts in piny on-juniper woodlands in Grand Canyon National Park. Am. Midi. Nat., 127:139-148. Bolger, D.J.. J.C. Pires, and J. Francisco-Ortega. 2006. Phylogeny of Agavaceae based on ndh F, rbcL, and its sequences: implications of molecular data for classification. Aliso. 22:313-328. Borchert, M. and C.M. Tyler. 2009. Patterns of post-fire flowering and fruiting in Chlorogalum pomeridianum var. pomeridianum (DC.) Kunth in southern California chaparral. Int. .1. Wildland Fire. 18:623-630. Brooks, M L., C.M. D'Antonio, D.M. Richardson, J.B. Grace, J.E. Keeley. J.M. DiTomaso, R.J. Hobbs, M. Pellant, and D. Pyke. 2004. Effects of invasive alien plants on fire regimes. BioScience, 54: 677-688. California Department of Fish and Game. 2007. California Natural Diversity Database. Calif. Dept. Fish Game, Sacramento. — . 2012a. Special vascular plants, bryophytes, and lichens list. Calif. Dept. Fish Game, Sacramento. 72 pp. — . 2012b. State and federally listed endangered, threatened, and rare plants of California. Calif. Dept. Fish Game, Sacramento. 16 pp. DiTomaso, J.M. 2000. Invasive weeds in rangelands: species, impacts, and management. Weed Sci, 48: 255-265. Eldridge, D.J. and R.S.B. Greene. 1994. Microbiotic soil crusts: a review of their roles in soil and ecological processes in the rangelands of Australia. Aust. J. Soil Res., 32:389—415. Fleischner, T.L. 1994. Ecological costs of livestock grazing in western North America. Conserv. Biol., 8: 629-644. Germano, D.J., G.B. Rathbun, and L.R. Saslaw. 2001. Managing exotic grasses and conserving declining species. Wildl. Soc. Bull., 29:551-559. Guretzky, J.A., E.R. Clark, and D. Woodbury. 2005. Life history traits of the threatened purple amole (Chlorogalum pwpureum var. purpureum ): Fort Hunter Liggett, California. ERDC/CERL Tech. Rep.-05-32. U.S. Army Corps Eng.. Champaign, Illinois. 77 pp. Hoover, R.F. 1940. A monograph of the genus Chlorogalum. Madrono, 5:137-147. . 1964. Some noteworthy Californian Liliaceae. Leaflets West. Bot.. 10:121-128. 1LTCN. 2001. IUCN Red List Categories and Criteria. Version 3.1. IUCN. Gland, Switzerland. 30 pp. — . 2010. Guidelines for Using the IUCN Red List Categories and Criteria. Version 8.1 (August 2010). IUCN, Gland, Switzerland. 85 pp. Jernstedt, J.A. 2012. Chlorogalum soap plant, amole. Pp. 1284—1285 In: The Jepson Manual: Vascular Plants of California (2nd ed.). (Baldwin, B.G.. D.H. Goldman, D.J. Kell. R. Patterson. T.J. Rosatti, and D.H. Wilken, eds.) Univ. Calif. Press, Berkeley. 1.600 pp. Klinger, R.C., M.L. Brooks, and J.M. Randall. 2006. Fire and invasive plant species. Chapter 22, Pp. 499-519 In: Fire in California's Ecosystems. (Sugihara, N.G.. J.W. Van Wagtendonk, K.E. Shaffer, J. Fites-Kaufman, and A.E. Thode, eds.) Univ. Calif. Press. Berkeley. 612 pp. Koch, A. and D. Hillyard. 2009. Camatta Canyon amole (Chlorogalum purpureum var. reduction) experimental restoration. P. 86 In: Abstracts for Oral and Poster Presentations: CNPS 2009 Conservation Conference-Strategies and Solutions. Calif. Native Plant Soc., Sacramento. 197 pp. Kucera, T.E. and K.E. Mayer. 1999. A Sportsman’s Guide to Improving Deer Habitat in California. Calif. Dept. Fish Game, Sacramento. 95 pp. Lesica, P. and J.S. Shelly. 1992. Effects of cryptogamic soil crust on the population dynamics of Arabis fecunda (Brassicaceae). Am. Midi. Nat., 128:53-60. Magney, D. 1988. Results of second year population dynamics of the state-listed rare Camatta amole (Chlorogalum purpureum var. reduction). Report to Los Padres National Forest, Goleta, Calif. 23 pp. National Park Service. 2007. Final Fort Hunter Liggett Special Resource Study. Natl. Park Ser., Oakland, Calif. 188 pp. Natural Resources Conservation Service. 2003. Soil Survey of San Luis Obispo County, California, Carrizo Plain Area. Nat. Resour. Conserv. Ser., Sacramento, Calif. 608 pp. Niceswanger, J. 2002. Purple amole (Chlorogalum purpureum var. purpureum) monitoring report: results summarized for monitoring 1998, 1999 and 2000. Report to Fort Hunter Liggett, Monterey County, Calif. 28 pp. 48 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Safford, H.D . J.H. Viers. and S.P. Harrison. 2005. Serpentine endemism in the California flora: a database of serpentine affinity. Madrono, 52:222-257. Stephenson. J.R. and G.M. Calcarone. 1999. Potentially vulnerable species: plants. Chapter 5, Pp. 223-316 In Southern California Mountains and Foothills Assessment: Habitat and Species Conservation Issues. USDA For. Ser. Pac. Southwest Res. Stn., Albany, Calif. 402 pp. U.S. Fish and Wildlife Service [USFWS]. 2000. Endangered and threatened wildlife and plants; determination of threatened status for Chlorogalum purpureum (purple amole), a plant from the south coast ranges of California. Fed. Register, 65:14878-14888. — . 2001. Endangered and threatened wildlife and plants; proposed designation of critical habitat for Chlorogalum purpureum , a plant from the south coast ranges of California. Fed. Register, 66: 56508-56531. — . 2002. Endangered and threatened wildlife and plants; final designation of critical habitat for Chlorogalum purpureum , a plant from the south coast ranges of California. Fed. Register, 67: 65414-65445. — . 2003. Endangered and threatened wildlife and plants; final designation of critical habitat for Chlorogalum purpureum , a plant from the south coast ranges of California; correction. Fed. Register, 68:20083 pp. — . 2008. Purple amole (Chlorogalum purpureum) five-year review: summary and evaluation. U.S. Fish Wildl. Ser., Ventura, Calif. 31 pp. White, P.J., W.H. Berry, J.J. Eliason, and M.T. Hanson. 2000. Catastrophic decrease in an isolated population of kit foxes. Southwest. Nat., 45:204-211. CONTENTS SMITHSONIAN INSTITUTION LIBRARIES I 1 111 I 111 II I 1 I t 19088 01 7( 3 6013 BHL Growth, Development, and Reproduction in Gulf Corvina (Cynoscion othonop- terus). Katie E. Gherard, Brad E. Erisman, Octavio Aburto-Oropeza, Kirsten Rowell, and Larry G. Allen 1 Recent Mass Mortality of Strongylocentrotus purpuratus (Echinodermata: Echinoi- dea) at Malibu and a Review of Purple Sea Urchin Kills Elsewhere in Califor- nia. Gordon Elendler 19 Review of the Purple Amole Chlorogalum purpureum (Agavaceae): a Threatened Plant in the Coast Ranges of Central California. Christopher P. Kofron, Connie Rutherford, Elizabeth R. Clark, Darlene Woodbury, Jody Olson, and Robert F. Holland 38 Cover: Sagittal otoliths of Gulf corvina. ISSN 0038-3872 & a ^ s&x SOUTHERN CALIFORNIA ACADEMY OF SCIENCES ll'llllllllllllllllllll<'l'l|>||l|||'HHi'l|||||lll|,.l,|ll|l|| l 9i see 9UllWU3d s» mm\n aivd jovisod so lldOUd NON WOW 30 N010NIHSVM HO l XX 83IUVUSI1 I S SZ HNWN H3X3 03V MN '3AV NOliniliSNOO ONV IS H10I NOliniliSNI NVIN0SH1IWS 2060S 30V Q3XIW A1 12(2) 49-160 (20 August 2013 Southern California Academy of Sciences Founded 6 November 1891, incorporated 17 May 1907 © Southern California Academy of Sciences, 2013 2012-2013 OFFICERS Julianne Kalman Passarelli, President Bengt Allen, Vice-President Edith Read, Recording Secretary Daniel Guthrie, Corresponding Secretary Ann Dalkey, Treasurer Daniel J. Pondella II, Editor Larry G. Allen, Editor ADVISORY COUNCIL Jonathan Baskin, Past President John Roberts, Past President Robert Grove, Past President John H. Dorsey, Past President Ralph Appy, Past President 2010-2013 BOARD OF DIRECTORS 2011-2014 Lisa Babilonia Brad R. Blood Ann Dalkey Julianne Passarelli Edith Read David Ginsberg Gordon Hendler Andrea Murray Dan Guthrie Gloria Takahashi 2012-2015 Bengt Allen Shelly Moore Ann Bull Dan Cooper Mark Helvey Membership is open to scholars in the fields of natural and social sciences, and to any person interested in the advancement of science. Dues for membership, changes of address, and requests for missing numbers lost in shipment should be addressed to: Southern California Academy of Sciences, the Natural History Museum of Los Angeles County, Exposition Park, Los Angeles, California 90007-4000. Professional Members $60.00 Student Members 30.00 Memberships in other categories are available on request. Fellows: Elected by the Board of Directors for meritorious services. The Bulletin is published three times each year by the Academy. Submissions of manuscripts for publication and associated guidelines is at SCASBULLETIN.ORG. All other communications should be addressed to the Southern California Academy of Sciences in care of the Natural History Museum of Los Angeles County, Exposition Park, Los Angeles, California 90007-4000. Date of this issue 19 September 2013 © This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Bull. Southern California Acad. Sci. 112(2), 2013, pp. 49-62 © Southern California Academy of Sciences, 2013 Developmental Mode in Opisthobranch Molluscs from the Northeast Pacific Ocean: Additional Species from Southern California and Supplemental Data Jeffrey H.R. Goddard' and Brenna Green2 ' Marine Science Institute, University of California, Santa Barbara, CA 93106-6150, goddard@lifesci. ucsb. edu 2 Department of Invertebrate Zoology and Geology, California Academy of Sciences, San Francisco, CA 94118-4503 Abstract. — We document development type for 33 species of benthic opisthobranch gastropods - 15 for the first time - collected mainly from the Southern California Bight. Fourteen of the newly examined species had planktotrophic development, while the dorid nudibranch Atagema alba had capsular metamorphic development, the first example of direct development in a non-dendrodoridid nudibranch known from the northeast Pacific Ocean. For the remaining 18 species our new data are either consistent with earlier determinations of development type, or confirm previous inferences. The new data also broaden geographic coverage for some species, and for the sacoglossan Stiliger fuscovittatus and the nudibranch Melibe leonina , suggest that egg size is inversely related to temperature. We correct the previous erroneous identification of nephrocysts as eyespots in the hatching planktotrophic larvae of the nudibranchs Tritonia f estiva and Janolus fuscus. These results further highlight the predominance of planktotrophic development in benthic opisthobranchs from the northeast Pacific Ocean. Introduction Mode of development has been determined for approximately two-thirds of the over 300 species of opisthobranch molluscs known from the northeast Pacific Ocean (Goddard 2004, 2005; Goddard and Hermosillo 2008). Small eggs and planktotrophic development predominate among these species; so far, only 14 species from the region are known to hatch from their egg coverings as either lecithotrophic larvae or juveniles (Goddard and Hermosillo 2008). The four species of nudibranchs known to bypass a free-living larval stage and hatch as juveniles all belong to a single family, the Dendrodorididae (Goddard, 2005). Developmental mode in opisthobranchs is most reliably documented when based on observations of the morphology of hatching veliger larvae - particularly the presence or absence of eyespots and propodium - accompanied by measurements of egg size, embryonic period and temperature, and size and type of shell at hatching (Bonar 1978; Goddard 2004). However, these parameters are not always readily measured or reported in the literature, and developmental mode has previously been inferred for some species based on subsets of the above information. For example, Goddard (2004) characterized as planktotrophic the development of the nudibranchs Polycera tricolor Robilliard, 1971 and Dirona picta MacFarland in Cockerell & Eliot, 1905, based on, respectively, (1) a comparison of the embryonic period reported for P. tricolor by Robilliard (1971) with those of congeners of known developmental type, and (2), qualitative observations, combined with a measurement of the preserved embryos reported by Marcus and Marcus (1967). As 49 50 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Table 1. Collection localities. Site GPS coordinates Depth (m)/habitat Whittier, Alaska Monterey Bay Monterey Harbor Asilomar, Pacific Grove Sand Dollar Beach, Monterey Co. Cayucos Hazard Canyon, Montana de Oro State Park Naples, Santa Barbara Co. Naples Reef Santa Barbara Harbor Tarpits Reef, Carpinteria Laguna Beach La Jolla South Casa Reef, La Jolla Bird Rock, La Jolla San Clemente Island Mission Bay, San Diego Point Loma, San Diego Punta Rosarito, Baja California El Tomatal, Baja California 60.7778, -148.6903 Floating docks 36.6088, -121.8797 Subtidal, 16 m 36.6043, -121.8912 5 m, dock pilings 36.6272, -121.9408 Rocky intertidal 35.9217, -121.4717 Rocky intertidal 35.4478, -120.9100 Rocky intertidal 35.2897, -120.8839 Rocky intertidal 34.4339, -119.9514 Rocky intertidal 34.4218, -119.9523 Subtidal, 15 m 34.4067, -1 19.6892 Floating docks 34.3869, -119.5164 Rocky intertidal 33.5429, -1 17.7906 Rocky intertidal 32.8549, -117.2680 subtidal, 17 m 32.8434, -1 17.2814 Rocky intertidal 32.8144, -117.2739 Rocky intertidal 32.8134, -118.3626 subtidal, 20 m 32.7642, -117.2172 Floating docks 32.6664, -117.2450 Rocky intertidal 28.5672, -114.1597 Rocky intertidal 28.4869, -114.0694 Rocky intertidal opportunities have arisen, we have endeavored to fill some of these gaps in the available embryological data, as well as sample across more of the geographic range of some species. Aside from providing confirmation or not of previous inferences, this will enable more robust comparative studies, help distinguish closely related species, and increase our understanding of geographic variation in the life history traits of opisthobranchs. Here we document mode of development for species not previously examined, most of which are from southern California, and fill existing gaps in the embryological data for others. We include species rarely observed or mainly subtidal in distribution, two undescribed species, and document a new example of rare direct development in a nudibranch from the northeast Pacific Ocean. We discuss discrepancies between our observations and those reported in the literature and also correct the previous erroneous identification by the senior author of the paired larval structures known as nephrocysts (e.g., Thompson 1976; Bonar 1978) as eyespots in the hatching larvae of two common planktotrophic species. Collection Sites and Methods Adult opisthobranchs, or portions of their egg masses in the process of being deposited, were collected by hand from subtidal and intertidal sites along the Pacific coast of North America (Table 1). Adults were held in containers (250 to 1000 ml) of unfiltered seawater at near ambient ocean temperatures until they laid egg masses. Recently laid egg masses were examined using a compound microscope equipped with an ocular micrometer. If first cleavage had not commenced, the diameters of a random sample of 10 zygotes were measured in each egg mass; otherwise, an upper limit on zygote size was estimated by measuring the dimensions of a few randomly selected embryos at or before the gastrula stage. We then isolated individual egg masses, or approximately I cm long sections of larger egg masses, into separate, labeled vials (20 ml) and changed and gently swirled the seawater in these once or twice daily. We examined the egg masses daily until hatching DEVELOPMENTAL MODE IN OPISTHOBRANCHS FROM SOUTHERN CALIFORNIA 51 and then measured the longest dimension of their shells and assigned mode of development (planktotrophic, lecithotrophic or direct) and larval shell type (coiled type 1 or egg-shaped, inflated type 2) as in Goddard (2004). After obtaining the above egg masses, adult specimens were relaxed in 7.5% MgCL, fixed in 70% ethanol, and deposited as voucher specimens in the California Academy of Sciences. Some of these vouchers included pieces of egg masses and were supplemented by digital images of adults. We used an underwater data logger (StowAway Tidbit, Onset Computer Corp.) to record temperature at 10 min. intervals in our holding containers. For the dorids Geitocloris mavis Marcus & Marcus, 1967 and Taringa aivica Marcus & Marcus, 1967, we examined preserved specimens originally collected in 1994 and 2000, respectively, and following Goddard (2004), inferred their mode of development based on the relatively small size of their embryos. For the chromodorid Felimare calif orniensis (= Hypselodoris calif orniensis) (Bergh, 1879) we inferred mode of development based on estimates of embryonic period and egg size. We estimated the latter by measuring the zygotes in an image of an adult laying its egg ribbon (Figure 2A) and assuming an egg ribbon width of 6 mm (adjusted to 4.243 mm to account for the approximately 45 viewing angle of the segment of the egg ribbon containing the zygotes we measured), based on the measurement reported by Ingram (1935). Craig Hoover (personal communication to JG, 30 Sept 2012) provided information on the embryonic period of F. calif orniensis, based on subtidal observations he made in September 2012 of three egg masses laid by F. calif orniensis in Big Fisherman Cove, Santa Catalina Island. JG identified the egg mass of Atagema alba (O'Donoghue, 1927) from Bird Rock, La Jolla (see Table 2) based on original 35 mm slides, including close-ups of sections of egg ribbon, in the James R. Lance collection at the California Academy of Sciences (CAS) in San Francisco. We follow the taxonomic nomenclature of Behrens and Hermosillo (2005), and for the Chromodoridae, Johnson and Gosliner (2012). Results We obtained results on the development of 33 species, 15 for the first time (Table 2). Eleven of these 15 laid small eggs and developed into hatching planktotrophic larvae (Table 2). Planktotrophic development was inferred for an additional three of them (Geitocloris mavis, Felimare calif orniensis, and Flabellina pricei), based on the small size of their embryos, and for Felimare calif orniensis, the length of its embryonic period as well (Table 2). Finally, Atagema alba from La Jolla developed from relatively large eggs and had capsular metamorphic development (Table 2, Figure 1 ). The juveniles of this species exited their shells while leaving the egg capsules, and one day after hatching had rhinophore buds and an oval notum reinforced by calcareous spicules and edged with secretory glands (Figure IB & C). Five days after hatching, the juveniles had two to three notal caryophyllidia (papillae surrounded by crowns of spicules; Figure ID). A radula with three pairs of hammate teeth was visible two days after hatching (Figure IF) and appeared to have approximately six rows of teeth after seven days (Figure IE). Of the remaining 18 species examined, hatching planktotrophic larvae were observed in 15, and this same mode of development was inferred for the remaining three based on the small size of their embryos (Table 2). Discussion Thirty-two of the 33 total species we examined were found to have planktotrophic development, consistent with prior compilations of developmental mode in benthic Table 2. New data on embryonic development of opisthobranch molluscs from California. Mode of development: P = planktotrophic, CM = capsular metamorphic, with modes inferred based solely on size of zygotes or early embryos in parentheses. An asterisk (*) preceding a species name indicates a species whose mode of deve- lopment has not previously been determined, and a dash ( — ) indicates no data. Values for egg diameter and shell length at hatching are means ± 1 SD, and sample size (n). 52 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES 2 Q O n ffl » 4 i/s — p cN tj “ On 3 3 § 2 §" 3 jft K N s '£ N N a ^ U < U 5 C/3 a U ex, cL .?2 CT £ £ O a? £ o o Embryonic Mode of Egg Diameter # eggs period Temp. Shell Shell length at Eyespots Develop- Taxon (pm) per capsule (days) (°C) type hatching (pm) at hatching ment Locality Palio dubia (M. Sars. 1829), <75d 1 — — — — (P) Whittier, Alaska CASIZ 182721 DEVELOPMENTAL MODE IN OP1STHOBRANCHS FROM SOUTHERN CALIFORNIA 53 C/3 i ^ o o o3 oo £ 3 6 o u, 03 u P o3 s o p +1 V + 1 + 1 1 + 1 + 1 + 1 +1 o p r-i p p H-. p oo r- ON r- © r- o' r- o r- NO NO 03 ■H S .s o oC S N Q C/3 ^ < •G u ^ r-~ 2 CN OO S M J3 co > < 2 U Cj ^ I b! « co -I < §> U P- ^ ■§ N = '' co Q 3 < ~g q o G O Q o a co < U N ±5 go 1 < | S L 5 'o js — • q SS P M c VO O U 15 2.^0 .2 o H H * s n oo N o ”0 G JS c/3 s < U- U o 03 i/D 5 O Os N 03 CO £ ^ g U u NO £ •so 1 «s .5 § a oo p> — < $ N o U Nudipleura: Aeolidina * Austraeolis steamsi (Cockerell, 81.4 ±2.2 (10) 1 7 18 1 122.0 ± 7.3 (10) no P La Jolla 1901). CASIZ 186486 *Babakinaf estiva 72.2 ±2.2 (10) 1 7.5 13-17 1 1 15.8 ± 1.1 (10) no P Naples (Roller, 1972), CASIZ 182204 Embryonic Mode of Egg Diameter # eggs period Temp. Shell Shell length at Eyespots Develop- Taxon (pm) per capsule (days) (°C) type hatching (pm) at hatching ment Locality Cuthona lagunae (O’Donoghue, 94.9 ± 1.1 (10) 1 8 — 2 262.1 ± 2.5(7) yes P Elazard Canyon 54 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES £ Oi _Q C3 C3 cZ "o CQ U C 3 03 O C/3 C/3 V V — ON C/5 N OO cJ c75 ON JO < in vo _ N — C/3 Os Shells left behind while hatching from egg capsules. Estimate based on image of egg ribbon (see Methods and Figure 2A). C. Hoover, personal communication to JG, 30 Sept 2012. Measurements of early embryos preserved in 70% ethanol. DEVELOPMENTAL MODE IN OP1STHOBRANCHS FROM SOUTHERN CALIFORNIA 55 Fig. 1. Live embryos and newly hatched juveniles of Atagema alba from Bird Rock, La Jolla. A. Embryos, still encapsulated, on day of hatching. B. Juvenile, left lateral view, one day after hatching. C. Juvenile, dorsal view, one day after hatching. I). Two juveniles, five days after hatching. E. Eyespot, with overlying lens, and radula from juvenile, 7 days after hatching. F. Radula from juvenile, two days after hatching. Note three pairs of hammate teeth. E and F from squashed specimens. Abbreviations used in figure: ec = egg capsule; ldg = left digestive gland; rhinophore = rhinophore bud. opisthobranchs from both the temperate and tropical northeast Pacific Ocean (Goddard 2004; Goddard and Hermosillo 2008). Atagema alba , with capsular metamorphic development, was the only species we observed in this study with non-feeding development. It represents the first example of direct development in a non- dendrodoridid nudibranch from the region. For each of the 18 species previously 56 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES D statocyst 100 pm Fig. 2. Nudibranch egg masses and veligers. A. Felimare californiensis laying an egg ribbon in the laboratory at the USC Marine Laboratory on Santa Catalina Island, September 1972. Image by John Engle. B. Newly hatched veliger larva of Eubranchus sp. 2 of Behrens and Hermosillo (2005), right lateral view. C. Near-hatching veligers of Tritonia festiva from Cape Arago, Oregon, July 1987. I). Near-hatching veligers of Janolus anulatus , showing (inset) higher magnification of a nephrocyst. examined, including Polycera tricolor and Dirornt picta , our new data either support earlier determinations of development type, or confirm previous inferences (see Goddard 2004, Table 1). Below we ( I ) discuss individual species, including discrepancies with previously reported embryological data, and (2) correct the previous erroneous identification by one of us (JG) of nephrocysts as eyespots in two species. Stiliger fuscovittatus Lance, 1962 Case (1972) studied this diminutive sacoglossan in San Francisco Bay and reported an uncleaved egg diameter of 70 pm and shell-size at hatching of 110 pm, similar to the values reported here for specimens from San Diego and Monterey. In contrast, Strathmann (1987) reported values of 95 pm and 150 pm for eggs and hatching larvae, respectively, from specimens from the San Juan Islands, Washington. A similar size discrepancy exists between our values for the eggs and larvae of Melibe leonina (Gould, 1852) from Santa Barbara (Table 2) and those reported by Strathmann (1987) for this species from the Washington and British Columbia. These size discrepancies are consistent with the inverse relationship between temperature and egg and larval sizes known for some marine invertebrates (discussed by Moran & McAlister 2009), and might be worth investigating experimentally. DEVELOPMENTAL MODE IN OP1STHOBRANCHS FROM SOUTHERN CALIFORNIA 57 Atagema alba (O’Donoghue, 1927) The egg mass of this species was found under an intertidal cobble at Bird Rock, La Jolla. The egg mass was a fairly stout ribbon 3 mm high, coiled in 1.3 turns, with a simple (not wavy) free edge and a total diameter of 6.5 mm. The zygotes and embryos were white. Compared to the images in the Lance collection at CAS of an egg mass laid in captivity by an A. alba collected intertidally in San Diego in July 1969, the egg mass we found had one less turn, but was otherwise virtually identical in appearance, including the relatively large size and spacing of the embryos. In the northeast Pacific Ocean, only Dendrodoris behrensi Millen & Bertsch, 2005 deposits a similar egg mass, but its embryos are significantly larger than those of A. alba , are deposited in thicker-walled capsules, and undergo ametamorphic direct development (Goddard 2005). Our voucher specimens (CASIZ 186493) of A. alba consist of post -metamorphic juveniles as pictured in Figure ID. Doris pickensi Marcus & Marcus, 1967 This species, previously known only from the northern Gulf of California to Costa Rica (Behrens and Flermosillo 2005; Camacho-Garcia et al. 2005), laid its egg ribbon flat, rather than on edge like most other dorids. The yellow ribbon, pictured in Goddard (2012a), measured up to 2.3 mm wide and was laid in a loose coil of three turns. Felimare calif orniensis (Bergh, 1879) Our estimate of 100 pm for the egg diameter of this species is based on Ingram’s (1935) measurement of a single egg ribbon, applied to Figure 2A, and therefore depends on variability in egg ribbon width in F. calif orniensis, which is unknown. However, the estimate is similar to the 95 pm egg size of another chromodorid, the sympatric Felimida macfarlandi (Cockerell, 1902), a known planktotroph (Goddard 2004), as well as planktotrophic representatives in the northeastern Pacific Ocean of the closely related genus Cadlina (Goddard 2004). Based on in situ observations by Craig Hoover of three egg masses in Big Fisherman Cove, Santa Catalina Island, F. calif orniensis developed to hatching at 17-19°C in at least seven days, but not more than 12 days (Table 2). At these temperatures, this range in embryonic period falls within that typical of dorid nudibranchs with similarly sized eggs and planktotrophic development and is also at least a week shorter than embryonic periods known for species with non-feeding modes of development in the NE Pacific Ocean (Hadfield and Switzer-Dunlap 1987; Goddard 1996; Goddard and Hermosillo 2008). We therefore conclude that F. calif orniensis has planktotrophic development. Based on an image in (Kopp 2008), a smaller congener of F. calif orniensis, F. porterae (Cockerell, 1901 ) also lays relatively small eggs and likely has planktotrophic development. Although the information we present here on the development of F. calif orniensis is incomplete and less precise than for other species, we have included it to help assess the cause of the loss of this species from the mainland of southern California. Felimare calif orniensis was once common throughout the Southern California Bight, but in contrast to other Californian nudibranchs in the region, has not been sighted on the mainland in over three decades (M. Miller in Behrens 2001; Goddard et al. 2013). Palio dubia (M. Sars, 1829) The size of the embryos from Alaska is consistent with measurements of the eggs and embryos of this species from the North Atlantic Ocean (Hamel et al. 2008; Goddard 201 la). 58 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Triopha maculata MacFarland, 1905 Mulliner (1972) reported that an 80 mm long Triopha grandis MacFarland 1966 “produced 400,000 eggs per egg-mass” and hatched after 6 days (at an unspecified temperature) into “free swimming, planktotrophic veligers.” Because both the large brood size and short hatching time are entirely consistent with planktotrophic development, Goddard (2004) followed Mulliner’s (1972) determination of develop- mental mode for this species, which was synonymized, along with Triopha occidentalis (Fewkes, 1889), with T. maculata by Ferreira (1977). Behrens (1991) and Goddard (2004) followed Ferreira (1977), but not McDonald (1983, 2007), who considers T. grandis a junior synonym of T. occidentalis , separate from T. maculata. Our finding that the eggs deposited by a 50 mm long T. maculata were encapsulated singly (Table 2) contrasts with Mulliner’s (1972) report of an average of 18 eggs per capsule for T. grandis. This large difference between egg masses in the number of eggs per capsule suggests that T. grandis may indeed be separate from T. maculata , as argued by McDonald (1983, 2007). Do to columbiana O’Donoghue, 1921 This species feeds on hydroids of the genus Aglaophenia and has previously been referred to by the senior author as Doto form B (Goddard, 1996, 2004). The sizes of the eggs and hatching larvae reported here from Hazard Canyon are similar to those reported from northern California by Goddard (1996). Janolus anulatus Camacho-Garcia & Gosliner, 2006 The cream to white egg strings of this species (see Goddard 2012b) contained embryos encapsulated singly. This contrasts with other species of Janolus known from the northeast Pacific Ocean, including J. barbarensis (Cooper, 1863), with 15 to 20 embryos per capsule (present study. Table 2), and J. fuscus O’Donoghue, 1924, with an average of 66 embryos per capsule (Wolf and Young 2012). The zygotes and hatching larvae of J. anulatus were also significantly smaller than those of the above two congeners (Goddard 2004; Wolf and Young 2012; present study, Table 2). Cumanotus sp. Following Behrens (1991), Goddard (2004) referred to this species, which feeds on Ectopleura crocea (Agassiz, 1862) in bays and harbors, as Cumanotus fernaldi Thompson & Brown, 1984. However, as pointed out by S. Millen (personal communication to JG, 8 Jan 2012), Thompson and Brown (1984) in their brief description of C. fernaldi were referring to the larger, soft-sediment dwelling species studied by Hurst (1967) that is ecologically and morphologically similar to the north Atlantic C. beaumonti (Eliot, 1906). Hurst (1967) used the name C. beaumonti for her specimens from Washington and reported that they laid corkscrew shaped egg masses with 4-14 eggs per egg capsule. Cumanotus sp. consistently has one egg per capsule (Goddard 1992; present study. Table 2), is ecologically and morphologically similar to the north Atlantic C. cuenoti Pruvot-Fol, 1948, and is undescribed (S. Millen, personal communication to JG, 8 Jan 2012). It is pictured as C. fernaldi in Behrens and Hermosillo (2005, species number 244), but that name actually applies to species number 245 in Behrens and Hermosillo (2005). (S. Millen, personal communication to JG, 8 Jan 2012). Unaware that two species of Cumanotus exist in the northeast Pacific Ocean, Goddard (2004, Table 1) lumped development data for both species. The complete and correct DEVELOPMENTAL MODE IN OPISTHOBRANCHS FROM SOUTHERN CALIFORNIA 59 breakdown is as follows. Based on Hurst (1967), C. fernaldi (as C. beaumonti ) deposits corkscrew-shaped egg masses with 4 14 eggs per capsule that hatch after 10 days at 8- 11°C with type 1 shells averaging 1 19 pm long. Based on Goddard (1992, 2011b) and the present study, Cumanotus sp. also deposits corkscrew-shaped egg masses, but with one egg (averaging 73 pm in diameter) per capsule, hatching after 9-10 days at 12-16 C with type 1 shells averaging 1 30 pm long. Cuthona lagunae (O'Donoghue, 1926) The egg masses and embryonic development of specimens of C. lagunae collected from Hazard Canyon in central California were virtually identical to those described by Goddard (1991) for specimens from northern California and southern Oregon, including the extrusion from the zygotes of unusually large, yolk-filled polar bodies. Eubranchus sp. 2 of Behrens and Hermosillo (2005) This species laid small, C-shaped egg masses characteristic of the genus. The embryos developed into planktotrophic veligers with type 2 shells, a small foot, eyespots, and minimal yolk reserves (Figure 2B). Flabellina goddardi Gosliner, 2010 The developmental data for this species in Table 2 were obtained from the coiled and secondarily looped, white egg string and near-hatching veligers pictured without size or scale information in Gosliner’s (2010) original description of this distinctive species. Since finding the first specimen in May 2008, we have found 14 additional specimens of F. goddardi , all in May and June at Tarpits Reef in Carpinteria, California, the type locality of this species. Flabellina trilineata (O’Donoghue, 1921) Two specimens found feeding on Ectopleura crocea in the Santa Barbara Harbor in January 2012 each laid egg strings with most capsules containing 2 eggs (Table 2). Previously, only one egg per capsule has been recorded for this species, including for specimens from the outer coast of Santa Barbara County (Bridges and Blake 1972; Strathmann 1987; Goddard 1992; present study). Flabellina sp. Adults were collected intertidally at Naples, on the south coast of Santa Barbara County. They closely resembled F. trilineata collected from the same locality in overall shape, but had smooth to slightly wrinkled rhinophores, more irregular white lines on the body, and white cnidosacs. Instead of a thin white line down the middle of the notum as in F. trilineata , some of our specimens of Flabellina sp. had a notum mostly covered with white pigment. Flabellina sp. deposited egg masses similar to those of F. trilineata collected from the same locality, but the ribbons of Flabellina sp. had a smaller diameter. The eggs were similar in diameter to those of F. trilineata , but the shells of the hatching veligers were slightly larger (Table 2). Preliminary genetic evidence suggests Flabellina sp. is distinct from F. trilineata (R. Johnson, personal communication to JG, 23 June 2010). In his description of Coryphella piunca, Marcus (1961, Fig. 163) illustrated smooth to lamellate rhinophores. Steinberg (1963) synonymized C. piunca with C. trilineata O’Donoghue, 1921, and later. Roller (1970) synonymized C. fisheri MacFarland, 1966 with C. trilineata. Because both MacFarland’s (1966) description of C. fisheri and 60 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES O’Donoghue’s description of C. trilineata only mention annulate or lamellate rhinophores, it appears that Marcus’s C. piunca likely included specimens of both F. trilineata and Flabellina sp. as recognized here. Goddard (1984, Table 1) reported that dendronotid Tritonia festiva has eyespots at hatching, and Strathmann (1987, p. 291 ) and Goddard (2001b, Appendix A) repeated this. Reexamination of JG’s original notes from 1981 and photomicrographs taken by JG in 1987 of the hatching larvae of this species revealed the paired structures were not eyespots, but rather nephrocysts (see Thompson 1976; Bonar 1978; Bickell and Kempf 1983) containing unusually large and dense, irregularly shaped, semi-translucent granules. At low magnification and certain angles of view and illumination, these structures can appear to be eyespots (Figure 2C). However, larval eyespots in opisthobranchs consist of thin patches of opaque black pigment, develop an overlying hyaline lens, and are located anterior to the nephrocysts (Thompson 1976; Bonar 1978). Goddard (1992, 2001b) also mistook dense nephrocysts for eyespots in the arminid nudibranch Janolus fuscus, but these have also been confirmed to be nephrocysts in both this species (Wolf and Young 2012; pers. obs.), as well as its congener, J. anulatus (present study. Figure 2D). The combined presence of eyespots and a propodium is a reliable indicator of metamorphic competence in opisthobranchs (Bonar 1978), and most planktotrophic nudibranchs hatch without eyespots (Bonar 1978; Hadfield and Switzer-Dunlap 1984; Goddard 2004). However, exceptions are known from the northeast Pacific Ocean, including species of Eubranchidae and Tergipedidae (Hurst 1967; Strathmann 1987; Goddard 1991 ; present study. Figure 2B) and the dorid Aegires albopunctatus MacFarland, 1905, which has unusual larval development (Goddard 2001a). Acknowledgements BG thanks the captain and crew of the Truth Aquatics dive boat Vision, as well as John Yasaki, Brynn Hooton, Terry Strait, Kim Mitchell and Brian Swift. We also thank Jack Engle for permission to use his image of Felimare calif orniensis, Maya Wolf for sharing with us her fine images of the hatching larvae of Janolus fuscus , and Craig Hoover for sharing his in situ observations on the egg masses of Felimare calif orniensis. Two anonymous reviewers commented on the manuscript, and we are thankful for their efforts. Some of the specimens documented in this study were collected during a study of intertidal nudibranchs funded by grant R/OPCENV-08 from the California Ocean Protection Council and California Sea Grant to JHRG. References Behrens, D.W. 1991. Pacific coast nudibranchs. Sea Challengers: Monterey, Calif., vi+107 pp. - . 2001. Hypselodoris californiensis (Bergh, 1879). Available from: http://slugsite.us/bow/nudwk255. htm. Accessed 15 Jan 2012. - and A. Hermosillo. 2005. Eastern Pacific nudibranchs. Sea Challengers, Monterey, Calif. vi+137 pp. Bickell. L.R and S.C. Kempf. 1983. Larval and metamorphic morphogenesis in the nudibranch Melibe leonina (Mollusca: Opisthobranchia). Biol. Bull., 165:119-138. Bonar, D.B. 1978. Morphogenesis at metamorphosis in opisthobranch molluscs. Pp. 177-196 in F.S. Chia and M.E. Rice (eds.). Settlement and metamorphosis of marine invertebrate larvae. Elsevier: New York, xi+290 pp. Bridges, C. and J.A. Blake. 1972. Embryology and larval development of Coryphella trilineata O'Donoghue, 1921 (Gastropoda: Nudibranchia). Veliger, 14:293-297. Camacho-Garcia, Y., T.M. Gosliner, and A. Valdes. 2005. Field guide to the sea slugs of the tropical eastern Pacific. California Academy of Sciences: San Francisco, Calif., 129 pp. DEVELOPMENTAL MODE IN OPISTHOBRANCHS FROM SOUTHERN CALIFORNIA 61 Case, R.M. 1972. An ecological study of Stiliger fuscovittata (Mollusca: Opisthobranchia). M.A. Thesis, California State University, San Francisco, 76 pp. Ferreira, A.J. 1977. A review of the genus Triopha (Mollusca: Nudibranchia). Veliger, 19:387^-02. Goddard, J.H.R. 1984. The opisthobranchs of Cape Arago, Oregon, with notes on their biology and a summary of benthic opisthobranchs known from Oregon. Veliger, 27:143-163. . 1991. Unusually large polar bodies in an aeolid nudibranch: a novel mechanism for producing extra-embryonic yolk reserves. J. Moll. Stud., 57(4)(Suppl.): 143-152. . 1992. Patterns of development in nudibranch molluscs from the northeast Pacific Ocean, with regional comparisons. Ph.D. thesis. University of Oregon, Eugene, Oregon., 237 pp. . 1996. Lecithotrophic development in Doto amyra (Nudibranchia: Dendronotacea), with a review of developmental mode in the genus. Veliger, 39:43-54. — . 2001a. The early veliger larvae of Aegires albopunctatus (Nudibranchia: Notodorididae), with morphological comparisons to members of the Notaspidea. Veliger, 44:398-400. — . Mollusca: Gastropoda. Pp. 86-128 in A.L. Shanks (ed.). An identification guide to the larval marine invertebrates of the Pacific Northwest. Oregon State University Press: Corvallis, Oregon, 314 pp. — . 2004. Developmental mode in opisthobranch molluscs from the northeast Pacific Ocean: feeding in a sea of plenty. Can. J. Zook, 82:1954—1968. — . 2005. Ametamorphic direct development in Dendrodoris behrensi (Nudibranchia: Dendrodor- ididae), with a review of developmental mode in the family. Proceed. Calif. Acad. Sci., 56:201-21 1. . 2011a. Polio dubia (Nudibranchia: Doridina) from the north-west Atlantic Ocean: is its morphology at hatching consistent with settlement one day later? J. Mar. Biol. Assoc. U.K.. 91: 1651-1654. — . 2011b. Cumanotus feeding on Ectopleura crocea. Available from: http://www.llickr.com/photos/ 34486353@N07/659225987 1 /in/set-72 1 576 1 278484 1131. Accessed 18 Dec 2012. — . 2012a. Doris pickensi laying eggs. Available from: http://www.fiickr.eom/photos/34486353@N07/ 6761 047 1 55/in/set-72 15761 278484 1131/. Accessed 1 Dec 2012. — . 2012b. Bryozoan colony after 4 days of predation by Janolus anulatus. Available from: http://www. fiickr.eom/photos/34486353@N07/7217848426/in/set-72157612784841 131. Accessed 1 Dec 2012. — and A. Hermosillo. 2008 Developmental mode in opisthobranch molluscs from the tropical Eastern Pacific Ocean. Veliger, 50:83-96. — , M.C. Schaefer. C. Hoover, and A. Valdes. 2013. Regional extinction of a conspicuous dorid nudibranch (Mollusca: Gastropoda) in California. Marine Biology doi: 10. 1 007/s00227-0 1 3-2204-x. Gosliner, T.M. 2010. Two new species of nudibranch mollusks from the coast of California. Proceed. Calif. Acad. Sci., 61:623-631. Hadfield, M.G. and M. Switzer-Dunlap. 1984. Opisthobranchs. Pp. 209-350 in A.S. Tompa, N.H. Verdonk, and J.A.M. van den Biggelaar (eds.). The Mollusca. Vol. 7. Reproduction. Academic Press, New York. Hamel, J.-F., P. Sargent, and A. Mercier. 2008. Diet, reproduction, settlement and growth of Polio dubia (Nudibranchia: Polyceridae) in the north-west Atlantic. J. Mar. Biol. Assoc. U.K., 88:365-374. Hurst, A. 1967. The egg masses and veligers of thirty northeast Pacific opisthobranchs. Veliger. 9:255-288. Ingram, W.M. 1935. Notes on egg masses associated with eight nudibranchs from Laguna Beach, California. J. Entom. Zook, Pomona College, 27:48-49. Johnson. R.F. and T.M. Gosliner. 2012. Traditional taxonomic groupings mask evolutionary history: a molecular phylogeny and new classification of the chromodorid nudibranchs. Plos ONE, 7:e33479. Kopp, K. 2008. July Wrinkles Dive on Catalina - Part 1 of 3 (the Macros). Available from: http://www. scubaboard.com/forums/socal/243321-july-wrinkles-dive-catalina-part-l-3-macros.htmk Accessed 26 Nov 2012. MacFarland, F.M. 1966. Studies of the Opisthobranchiate Mollusks of the Pacific coast of North America. Memoirs of the California Academy of Sciences, 6:1 546. Marcus, E. 1961. Opisthobranch mollusks from California. Veliger, 3(Supplement): 1-85. Marcus, Er. and Ev. Marcus. 1967. American opisthobranch mollusks. Stud. Trop. Oceanog. (Miami), 6: T 256. McDonald, G.R. 1983. A review of the nudibranchs of the California coast. Malacologia, 24:114-276. — . 2007. Sacoglossa and Nudibranchia. Pp. 788-807 in J.T. Carlton (ed.), The Light and Smith manual: intertidal invertebrates from central California to Oregon. University of California Press: Berkeley, Calif, xvii+1001 pp. 62 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Moran, A.L. and J.S. McAlister. 2009. Egg size as a life history character of marine invertebrates: is it all it’s cracked up to be? Biol. Bull., 216:226-242. Mulliner, D.K. 1972. Breeding habits and life cycles of three species of nudibranchs from the eastern Pacific. Festivus, 3:1-5. Robilliard, G.A. 1971. A new species of Polycera (Opisthobranchia: Mollusca) from the northeastern Pacific, with notes on other species. Sysesis, 4:235-243. Roller, R.A. 1970. A list of recommended nomenclatural changes for MacFarland’s “Studies of opisthobranchiate mollusks of the Pacific coast of North America.” Veliger, 12:371-374. Steinberg, J.E. 1963. Notes on the opisthobranchs of the west coast of North America - III. Further nomenclatorial changes in the order Nudibranchia. Veliger, 6:63-67. Strathmann, M.F. 1987. Reproduction and development of marine invertebrates of the northern Pacific coast. University of Washington Press: Seattle, xii+670 pp. Thompson, T.E. 1976. Biology of opisthobranch molluscs. Vol. I. Ray Society: London. 207 pp. and G.H. Brown. 1984. Biology of opisthobranch molluscs. Vol. II. Ray Society: London. 229 pp. Wolf, M. and C.M. Young. 2012. Complete development of the Northeast Pacific arminacean nudibranch Janolus fuscus. Biol. Bull., 222:137-149. Bull. Southern California Acad. Sci. 112(2), 2013, pp. 63-73 © Southern California Academy of Sciences, 2013 Small Mammal Use of the Burn Perimeter Following a Chaparral Wildfire in Southern California Mark Borchert1 and Sinead M. Borchert2 'San Bernardino National Forest, P.O. Box 292, Fawnskin, CA 92333, borchert l@charter. net 2Scliool of Renewable Natural Resources, Louisiana State University, Baton Rouge, LA 70803 Abstract. — Wildfires in southern California chaparral burn at high intensities and often cover thousands of hectares. Some small mammals survive the fire, while others colonize from scattered unburned islands and from intact vegetation bordering the main fire perimeter. For ten years (2002-201 1 ) we live-trapped two grids and used the number of captures to examine post-fire small mammal use of a narrow 65-m zone straddling the high-contrast edge between burned and unburned chaparral on the perimeter of a high-intensity wildfire. Results indicate that agile kangaroo rats ( Dipodomys agilis ) were captured more often in open, burned areas than in unburned chaparral. Deer mice (Peromyscus maniculatus) were captured equally in burned and unburned chaparral but did not show an affinity for either habitat or the edge of the burn. Pinyon mice (Peromyscus truei) were captured most often in unburned chaparral throughout the study but were prevalent on the burn edge in years one and four. In the first year post-fire, California mice (Peromyscus californicus ) were captured more frequently in unburned than burned chaparral but in years four and five, captures shifted toward the edge and then into the bum areas in year nine. We did not find evidence that any of the four species were dedicated edge specialists in this study. Neither pinyon mice nor California mice appeared to be permanent residents of the burns in the first ten years post-fire. We suggest that future research on post-fire small mammal succession in chaparral would benefit from chronosequence studies that give a more comprehensive, long term picture of succession. Introduction Chaparral wildfires in southern California, especially those driven by hot, dry Santa Ana winds in the fall, spread quickly, burn at high intensities, and have the potential to reach thousands of hectares in size. Post-fire temporal and spatial patterns of burn-area re-colonization by small mammals depend on a several factors, but primarily on the number of animals surviving the fire in situ (Banks et ah, 2011) and on animals re-entering the burn from refugia ( e.g ., unburned brush, rock outcrops, riparian zones). Since chaparral fires usually do not leave vegetation mosaics, unburned islands within large (thousands of hectares) burns often are few in number, vary greatly in size, and tend to be widely scattered (Quinn 1990; Borchert et ah, 2003). Compared to unburned islands, intact chaparral bordering the fire perimeter offers a more dependable source of colonists and is more likely to harbor the full contingent of small mammal species, and in greater numbers, than vegetation islands (Longland and Bateman, 2002). Management strategies for conserving faunal diversity after wildfires depend on a multi-scale understanding of fire effects (Di Stefano et ah, 2010). The earliest post-fire 63 64 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES small mammal studies in chaparral focused on the colonization of individual sites in burned and unburned chaparral (Lawrence 1966; Wirtz 1977; Wirtz et ah, 1988; Wirtz 1995). Often, however, proximity of burn trapping grids to the nearest unburned refugia was not stated and it was assumed that the rate of re-occupation was not dispersal-limited (Rosario and Mathias, 2007). More recently, research on small mammal succession has shifted to much larger scales (Schwilk and Keeley, 1998; van der Ree and Loyn, 2002; Brehme et ah, 201 1 ; Diffendorfer et ah, 2012). Trapping grids in these studies were placed in both unburned vegetation as well as in the burned area at varying distances from known refugia. Results of these studies have provided a more complete picture of the considerable variability in small mammal re-colonization that can occur at the landscape scale. A conspicuous feature of all chaparral wildfires is the high-contrast, hard edge along the burn perimeter and at the margins of unburned islands. While there is extensive literature on the effects of edges on small mammals in unburned habitats (e.g., Murcia 1995; Lidicker 1999), few studies have addressed small mammal use of fire-induced edges, even though edges play an important role in the recovery of the fauna in burned areas (Larrivee et ah, 2008; Santos et ah, 2009; Tasker et ah, 201 1). For example, we know of only three studies on small mammal use of fire-induced edges (Figueiredo and Fernandez, 2004; Pires et ah. 2005; Diffendorfer et ah, 2012). The habitat accommodation model (Fox 1982) has been proposed to explain small mammal succession in burned areas. This model relates small mammal re-colonization to post-fire changes in vegetation structure and composition, which also makes the model applicable to small mammal use of fire-caused edges. More recently, Ries and Sisk (2004) proposed a specific model to predict faunal distributions at edges of all types. Their model focuses on resource differences between the two juxtaposed habitats and the edge that separates them. In addition to creating sharp differences in vegetation, the edge separates habitats with markedly different food resources, risks of predation and abiotic factors. For example, in burned chaparral the early post-fire flora produces an abundance of seeds that may be missing, or occur at low densities, in unburned chaparral (Tyler and Borchert 2002). Moreover, as obligate seeding shrubs re-establish in the burn area, supplemental seeds may be abundant for several decades. On the other hand, unburned chaparral provides better protection from predators, especially in the early post-fire years. This asymmetry in resources creates conditions in which small mammals may increase their use of the edge because they can readily access an abundant food supply in the burn and at the same time find protection from aerial predators and carnivores in adjacent unburned chaparral. In this study, we focused on small mammal use of a narrow zone (65 m) straddling burned and unburned chaparral on the main fire perimeter of a wildfire. Based on the Ries and Sisk model, we expected small mammals to increase their activity at the burn edge, especially in the early post-fire years when the seed supply increased in the burn area. However, as the edge softened in later years, and burned vegetation became increasingly similar to the unburned chaparral, we expected the attraction to the edge to be less pronounced, or to disappear altogether. Therefore, our objectives were ( 1 ) to describe small mammal use of a high-contrast edge over a 10-year period, and (2) to compare edge and bum-area use to existing studies of small mammal succession in burned chaparral at greater distances from the burn perimeter. Methods Study Area In June 2002, the 8,100-ha Wolf Fire burned the two study sites situated on an extensive (~ 40 ha), old river terrace 30 m above Sespe Creek at 955 m elevation in Los SMALL MAMMAL USE OF A BURN PERIMETER 65 Padres National Forest. The study area is located 15 km NNE of Ojai, CA. The trapping grids were ~ 560 m NNW of the Sespe Trailhead (34° 33' 38" N, 1 19 9' 52" W). Before the fire, the study area was dominated by 70-year-old chamise (Adenostoma fasciculatum) chaparral, with widely scattered dead chaparral whitethorn ( Ceanothus leucodermis). Climate of the study area is Mediterranean consisting of cool wet winters and long dry summers. Average annual precipitation for Ojai, CA is 539 mm. From 2002-2010 precipitation was above average in five years and below average in four years. The fire burned at unusually high intensity through the entire study area but was extinguished on the fiat terrace ~ 120 m before reaching Sespe Creek. At the locations of the trapping grids, the boundary between burned and unburned chaparral was sharply defined because a 5-m wide bulldozer line was constructed as part of the fire suppression effort. On one side of the dozer line was a linear edge of unburned chaparral and on the burned side of the line a narrow (0.5 m-1.0 m) partially burned strip of vegetation left by the dozer we termed the “fringe”. Beyond the narrow fringe was the expansive, denuded burn area. In the early years after the fire the cleared dozer line was sparsely vegetated but by nine years chamise cover was —40%. Small Mammal Trapping We established two 8 x 12 trap grids 1 10 m apart. At each site, a row of 12 traps spaced 10 m apart was placed along the linear edge of the chaparral. Another row of traps was placed along the fringe five meters opposite the edge traps. Additional rows of traps were arrayed in the burn areas 15, 25 and 35 m perpendicular to the edge traps and 10, 20 and 30 m in the interior of the unburned chaparral. We used a single, large Sherman trap ( 10 cm X 12 cm X 38 cm) at each station. We opened traps in the afternoon the first day and trapped for five days (four trap nights). Traps were baited with sunflower seeds and pieces of peanuts and walnuts. Cotton balls served as bedding. We checked traps each morning. Captured animals were marked at the base of the tail with a permanent marking pen, weighed, sexed, and assessed for reproductive condition. We used weight and pelage color to distinguish adult from sub-adult animals. Trapping began in July 2002, one month after the fire. We trapped both grids in the spring (May-July) and fall (September-November) for the first five years post-fire (2002- 2006). At the end of five years, several species lacked a sustained presence in the burned areas so we delayed trapping until the spring of 2008 and again in the spring of 201 I to monitor changes in their status in the burn areas. Vegetation Methods For each grid we estimated the cover of shrubs, herbs, forbs and grasses to the nearest 5% within 7-nr (radius 1.5 m) plots using the trap station as plot center. Cover was sampled at burn stations one, three and eight years post-fire and at the unburned stations one and eight years post-fire. To measure the vertical stratification of foliage at all the stations, we randomly chose 1 nr within each 7-nr macro-plot. We first estimated the total shrub cover and height in the I -nr plot after which we placed a 2. 5-m pole alternately banded in 50 cm increments at the plot center. The estimated cover of the I -nr plot was partitioned among the 50-cm intervals marked on the pole. Rodent Abundance We used the Schnabel capture-recapture model to calculate the number of rodents per hectare in each trapping session. Densities are presented for the two grids combined 66 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Table 1. Vegetation percent cover in burned chaparral in 2003, 2005 and 2010 and in unburned chaparral in 2003 and 2010. Values are means ± 1 SE. 2003 2005 2010 Unburned Burned Burned Unburned Burned Herbs, forbs, and grasses 0 31.4 ± 2.8 14.0 ± 2.7 0 0 A. fasciculatum 71.2 ± 1.8 19.7 ± 2.7 35.8 ± 3.4 69.5 ± 1.9 42.9 ± 2.2 C. leucodermis 0 0 2.0 ± 3.4 0 16.9 ± 2.1 which summed to 1.43 ha. Because animals were batch-marked, densities could not be calculated separately for burned and unburned areas. Statistical Analysis We spread fourteen trapping sessions over nine years. Sessions were lumped to create yearly intervals for analysis. For example, sessions at 1 and 6 months were combined for the early post fire period (year 0). Yearly intervals with their corresponding sessions are as follows: year one (11 and 14 months), year two (18, 22 and 24 months), year three (34 and 40 months), year four (46 and 51 months), year five (58 months), year seven (85 months) and year nine (107 months). We used captures of each species to analyze the use of the four habitats over time: burned, fringe, edge and unburned. We analyzed habitat use by calculating the mean captures per trap (total captures in 4 nights) for 72 traps in the burned areas, 24 fringe traps, 24 edge traps and 72 traps in the unburned areas. Captures for the two grids were combined for the analyses since they were identical in rodent species composition and combining them increased the sample size. Because capture data were zero-inflated and failed to meet the assumptions of normality, we used Kruskal-Wallis analysis of variance to test for differences in captures per trap among the habitats for each of the eight years. RESULTS Vegetation Structure The high-intensity fire completely denuded the burn area. All that remained immediately after the fire were exposed chamise burls with short ( — 15 cm), charred stems. Post-fire annual cover was highest in 2003 and was composed almost entirely of the annual short-lobed phacelia (Phacelia brachyloba ) (Table 1). By 2005 the cover of annuals and short-lived perennials had all but disappeared in the burn. Living chaparral whitethorn was absent in unburned chaparral but after the fire, seedling production by this shrub was prolific so that by 2005 cover had reached 2% (Table 1). Between 2005 and 2010 the cover and height of chaparral whitethorn increased rapidly and by 2010 it comprised 17% of the burn area shrub cover (Table 1). Chamise both resprouted and produced seedlings after the fire; its height and cover increased steadily in the burned area. After eight years, average total shrub cover did not differ significantly between burned and unburned stations (69.5% burned vs. 59.8% unburned, tg2 = —1.07, P = 0.29) although shrubs still were taller in the unburned vs. the burned areas (178.5 cm unburned vs. 142.2 cm burned, tgi = 6.02, P < 0.001). The vertical foliage distribution was markedly different between burned and unburned vegetation eight years post-fire (Pig. 1). Compared to unburned chaparral, higher percentages of shrub foliage in the burned areas were in the 0-50 and 51-100 cm classes SMALL MAMMAL USE OF A BURN PERIMETER 67 CM E a> > o o +-> c a) o u.